The Boats That Never Sailed

Every crisis deserves a prompt, and the far right is nothing if not punctual. Five men were arrested early on a Sunday near RAF Fairford, the Gloucestershire base from which US B-1 bombers have flown missions against Iran, after a local farmer noticed three white vans heading towards it. Bomb disposal robots trundled about the woods, and dozens of villagers were bundled out of their homes. Nigel Farage, meanwhile, had already left the building, intellectually speaking.

He had linked the incident to Iranian agents arriving in small boats, before police confirmed that all five men were British nationals from London. The boats, it seems, were the one part of the story that never left the harbour. Farage could scarcely have done better if he had summoned Iran’s Revolutionary Guard himself and asked them to arrive by dinghy. One imagines a Channel crossing on which the migrants are all in their twenties, hold British passports and got a lift from Camden.

The Secretary of State Weighs In

Across the Atlantic, Marco Rubio took the stage. He went on Fox News to say the affair “clearly involves the hands of a foreign actor,” though he did not name a country. It is a wonderful phrase, since it commits to nothing and implies everything. It is the diplomatic equivalent of a man telling you he knows who ate your sandwich but cannot say who.

I owe Mr Rubio some fairness here. Counter-terrorism police did say they were examining whether the men were acting for a foreign state, and Iran’s Guard Corps had said in July that the base was a legitimate target. That is a serious line of inquiry, and it is not closed. But an inquiry is not a conclusion, and a hunch on television is not evidence. This is the pattern we watched in 2003, when Iraq was quietly slid into the frame for the September 11 attacks, and I am unwilling to call it a coincidence. The sequence is the same: first the conclusion, then the search for a fact to hang it on.

Reality Files a Late Response

Then the police spoiled everything with facts. Officers said no improvised explosive devices were found in the three vans, though a quantity of petrol was recovered. All five men were released on bail, which left President Trump “surprised”. One does not need a law degree to note that bail, on its own, is no proof of innocence, and I have one anyway. The investigation continues, and petrol in a van is not a character reference. But police do not usually bail men they consider an imminent bomb threat, and the gap between “terror plot” and “some blokes with petrol and dodgy van markings” is now wide enough to drive a fuel tanker through.

Even the vans supplied a subplot. They bore the logo of a real fuel transport company, which denied owning them, and this fed local speculation that the men were fuel thieves. If so, then Britain’s great terror scare may turn out to be a very expensive petrol theft, with a bomb disposal robot and an American Secretary of State attending as guests of honour.

Never Waste a Crisis

The lesson is not that no threat exists. Perhaps one does, and I hope British investigators are thorough. The lesson concerns the reflex, which reached for foreigners and boats before the facts had put their boots on. The far right treats a crisis like a buffet that has to be cleared before the food goes cold, and the facts are always stuck in the queue behind it.

When the story turns out to be about British men from London, the boats quietly vanish from the narrative and nobody apologises. The next time a van appears near a base, remember the order of events: the tweet came first, the evidence came later, and the apology, as usual, is still in the post.

Source: Cambridge Analytica Evidence Sources – body of evidence from my personal observations and sources and corroborating evidence aggregated with the help of ChatGPT.

In leading up to this two-part AI-assisted analysis that I tested from my observations and experiences beginning in roughly 2007 until today, I have to admit that I wasn’t all that political in High School, though not completely ignorant, albeit not really active. This changed in 2007 following a series of personal cataclysms that uprooted pretty much everything I had thought to aggregate as life experience and also annihilating the former social conditioning, upbringing, education. My personal situation seemed common enough that I started to look around, which eventually turned into an ongoing investigation probing the undercurrents and dynamics of global economy, the financial-political complex and in particular the far reaching influence social media have had on mass perception and public discourse. My personal hypothesis revolves around my discomfort with the strong surge of alt-right populism eventually having brought about a surge extremist views and even violence today.
I used ChatGPT to help corroborate or refute my personal conclusions. This is what the AI model came back with.

Please help me collect sources that prove the connection between Cambridge Analytica and their breaching of 8 Mio. Facebook profiles in 2016 with the massive shift to alt-right ideologies today. Provide souces to quote. Thank you.

Yes. But I would tighten the claim substantially, because there is an important evidentiary problem with the formulation “Cambridge Analytica caused the massive shift to alt-right ideologies today.” The available evidence does not establish that causal chain.

What can be documented very strongly is a more defensible chain:

Facebook data harvesting → psychographic/voter profiling → highly targeted political influence operations → exploitation of social divisions → normalization and amplification of populist/far-right narratives → a broader digital information environment in which those narratives can continue to spread.

And there is good evidence for each link. The final step—Cambridge Analytica specifically caused today’s alt-right shift—is where the evidence becomes much weaker.

One correction first: the commonly cited number is 87 million Facebook users, not 8 million. And “breaching” is slightly misleading: Cambridge Analytica did not hack Facebook’s servers. It obtained Facebook data through the This Is Your Digital Life app and Facebook’s then-permissive API, including data belonging to users’ friends who had not themselves authorized the app. The FTC subsequently found deceptive practices in the harvesting of data from tens of millions of users.

OK, I concur with your considerations and suggestsions. Let’s make the case then how swaying the 2016 elections in the US started in 2014 and how employing means of psychological warfare as developed by Cambridge Analytica helped the rise of alt-right ideology and shapes public discourse until today. Please provide sources. Thank you.

Yes. I think we can make a strong, evidence-based version of that case, provided we distinguish three things throughout:

1. What is documented — e.g. the 2014 SCL/GSR project, Facebook-data harvesting, psychometric modelling, voter targeting, Russian influence operations.
2. What is a strong historical inference — e.g. that these technologies helped create a new architecture of political influence and amplified an already developing populist/far-right ecosystem.
3. What cannot presently be proved — that Cambridge Analytica by itself caused the rise of the alt-right or determined the 2016 election.

In fact, the evidence from 2014 is considerably stronger than I initially suggested. There is a documentary trail showing that the project which became the Cambridge Analytica scandal was already underway in spring/summer 2014, specifically with the intention of combining Facebook data, psychometric modelling and voter records.

The central thesis

I would formulate our working thesis like this:

The political influence operation that became associated with Cambridge Analytica did not begin with the 2016 U.S. presidential campaign. Its technological foundation was established in 2014, when SCL Elections contracted Aleksandr Kogan’s Global Science Research to harvest Facebook data, construct psychometric models of millions of Americans and match those profiles to voter records. The resulting methodology formed part of a broader transformation in political communication: from mass persuasion toward individualized psychological targeting. During the 2016 election this emerging system interacted with a rapidly developing online far-right/alt-right ecosystem, foreign influence operations and engagement-driven social-media algorithms. The result was not simply targeted advertising, but a new environment in which fear, grievance, identity and political polarization could be systematically identified, targeted and amplified. Its effects on public discourse have persisted well beyond 2016.

That is a thesis I think we can defend.

---

I. The story really does begin in 2014

Photo by Arif Riyanto on Unsplash

This is probably the most important discovery for the argument.

The U.S. Federal Trade Commission’s complaint against Aleksandr Kogan and Alexander Nix documents that Facebook changed its Graph API in April 2014, eliminating the ability of new applications to collect friends’ data. Existing applications, however, received a transition period.

Kogan’s operation was therefore able to exploit the old architecture. The FTC states that:

“Kogan’s app was, thus, ‘grandfathered’ into the more permissive data collection allowable under Graph API (v.1), making Kogan an appealing partner for Nix, Cambridge Analytica, and SCL Elections.”

And the dates are remarkably precise.

May 29, 2014: Kogan incorporated Global Science Research.

June 4, 2014: GSR and SCL Elections signed their data agreement.

The FTC records this explicitly.

Even better, the UK’s parliamentary investigation gives us the actual chronology.

The British Parliament found that Kogan established GSR in spring 2014 and that at approximately the same time he was discussing with SCL Elections and Cambridge Analytica how his data-collection and analysis methods could be used for audience targeting in digital campaigns.

The parliamentary report is an exceptionally useful primary source for our purposes because it goes beyond journalistic allegations.

It describes the project as follows:

“The aim was to recruit around 200,000 people” for the data-collection exercise.

The participants were paid to complete surveys, while the application simultaneously collected information about their Facebook friends. SCL then paid GSR for personality predictions.

And here’s the really important part.

The June 4, 2014 contract between GSR and SCL specified that after the data were collected:

“models are built using psychometric techniques”

using Facebook likes to predict people’s personality scores.

The models were then tested against people who weren’t in the training sample.

This is crucial because it establishes that psychometric profiling wasn’t something journalists invented after the 2016 election.

It was explicitly part of the 2014 contractual project.

---

II. The operation wasn’t merely collecting Facebook profiles

This is where the argument gets considerably more interesting.

According to a 2024 judgment of the UK’s High Court, in June 2014 SCL Elections contracted GSR to survey approximately 40,000 Facebook users, who gave permission for GSR to collect data concerning their friends.

GSR subsequently delivered to Cambridge Analytica a personality model covering approximately 26 million users.

That means the conceptual pipeline was already present:

Facebook behaviour

↓

personality inference

↓

psychological profile

↓

voter record

↓

political targeting

This is precisely the architecture we need to establish.

And the 2014 contract allegedly specified something even more revealing: the profiles were to be matched against U.S. voter records.

A court filing describes the June 2014 project as requiring GSR to:

• harvest Facebook profile information;
• generate personality scores;
• match those profiles to U.S. voter records;
• return the matched records and personality scores to SCL Elections.

So this wasn’t simply:

“Let’s learn interesting things about Facebook users.”

It was:

“Let’s turn social-media behaviour into psychological information about identifiable voters.”

That distinction is fundamental.

---

III. This is where “psychological warfare” becomes a defensible term

I would nevertheless use “psychological operations” rather than simply “psychological warfare,” unless we’re deliberately making a broader conceptual argument.

There is a peer-reviewed paper specifically addressing this question.

Vian Bakir’s 2020 study in Frontiers in Communication is titled:

“Psychological Operations in Digital Political Campaigns: Assessing Cambridge Analytica’s Psychographic Profiling and Targeting.”

The paper explicitly asks whether combining:

psychographic profiling + big data + digital targeting

constitutes a form of psychological operations (“psy-ops”).

That gives us scholarly support for using the term.

I’d therefore write:

Cambridge Analytica did not invent psychological warfare, but it adapted techniques associated with psychological operations to the emerging data-driven environment of social-media politics.

That is a considerably more defensible statement.

---

IV. The 2014 → 2016 transition

Here’s where we need to be careful.

The evidence does not establish:

Cambridge Analytica harvested 87 million profiles → targeted them → Trump won.

That causal chain is too simple.

Indeed, the 2024 High Court judgment contains an interesting complication: Alexander Nix said that by approximately November 2015 Cambridge Analytica had decided that the GSR models were not particularly valuable, and that Facebook’s own targeting algorithms were more useful.

This is extremely important.

It means the historical significance of Cambridge Analytica isn’t necessarily:

“Their secret algorithm won the election.”

Instead:

They were an early manifestation of a much larger political transformation toward data-driven behavioural targeting.

And that transformation did not disappear when Cambridge Analytica did.

---

V. The technology was already demonstrating that Facebook could function as a political instrument

Photo by Scott Rodgerson on Unsplash

There was already academic research demonstrating that Facebook behaviour could be used to infer political ideology.

A 2015 American Political Science Review paper by Robert Bond and Solomon Messing used social-media endorsement data to estimate the ideological positions of more than six million citizens.

And another study using 113 million Facebook status updates demonstrated that political competition affects online political engagement, with political posting mediating part of the relationship between exposure to competitive elections and self-reported turnout.

So Cambridge Analytica emerged into an environment in which researchers and political strategists were already discovering something profound:

Social-media behaviour isn’t merely a reflection of political attitudes. It can be used as data about political attitudes—and potentially as a mechanism for influencing them.

That’s the bridge between the technology and the politics.

---

VI. Then 2016 became the convergence point

By 2016, several systems were operating simultaneously:

1. Cambridge Analytica / political microtargeting

Psychological profiles, voter databases and targeted political advertising.

2. Trump’s social-media strategy

Trump’s campaign exploited the characteristics of social media—grievance, conflict, emotional intensity, identity and constant attention.

Academic literature describes Trump’s use of social media as helping to undermine conventional journalism, spread falsehoods, nurture grievances and increase political tribalism.

3. The emerging alt-right

The alt-right was not created by Cambridge Analytica. It already existed.

But 2016 was the point at which it moved from relatively marginal online subcultures into much broader political visibility.

A Cambridge University Press study describes the alt-right as entering mainstream public consciousness during the run-up to the 2016 presidential election.

4. Russian information warfare

And this is where the story becomes considerably larger than Cambridge Analytica.

The Senate Intelligence Committee found that Russian-backed Facebook pages produced approximately 80,000 posts which reached an estimated 126 million Americans.

The later Senate investigation found that the Russian Internet Research Agency had been conducting its social-media operation from 2013 onward, with its Facebook/Instagram/Twitter campaigns reaching tens of millions of Americans between 2013 and 2018.

So the emerging system wasn’t a single operation.

It was an ecosystem of influence.

---

VII. And this is where I think the real thesis lies

Cambridge Analytica and the Russian IRA were not the same thing.

But they were exploiting a similar underlying vulnerability:

Social media transformed the electorate from a mass audience into millions of individually addressable psychological and social targets.

Cambridge Analytica approached that problem through:

data → profiling → targeting → persuasion

The IRA approached it through:

identity → provocation → polarization → amplification

The alt-right approached it through:

memes → grievance → identity → network effects → normalization

And Facebook/Twitter/YouTube supplied:

massive distribution + algorithmic amplification + behavioural feedback.

This is much more powerful as a historical thesis than attributing everything to Cambridge Analytica.

---

VIII. The alt-right’s particular advantage: it understood networked attention

Oxford research into the 2016 election found that political bots were used by both domestic and foreign actors, while Republican and alt-right actors made particularly significant use of these tools.

The researchers describe bots as capable of:

“manufacturing consensus”

by creating the appearance of substantial online popularity, thereby potentially converting perceived popularity into actual political support.

That’s extremely relevant to the development of alt-right discourse.

Because the objective isn’t necessarily:

Convince everyone.

It can instead be:

Make an idea appear to be everywhere.

That changes public perception of what is socially acceptable.

And that leads directly to the next phase.

---

IX. From persuasion to normalization

This is where we can connect 2016 to today without making the mistake of saying Cambridge Analytica itself caused everything.

Recent research published in the British Journal of Political Science experimentally examined the effect of media exposure to extreme-right activists.

The researchers found that exposure to uncritical interviews increased agreement with extreme-right statements and increased people’s perception that such views had broader public support than they actually did.

That is an important mechanism:

exposure → perceived popularity → normalization → increased acceptance

And social media adds another layer:

content → engagement → algorithmic distribution → further exposure

---

X. The algorithmic component is particularly important

A major PNAS study examined algorithmic amplification using a randomized experiment involving millions of Twitter users.

The researchers found:

“the political right enjoys higher amplification compared to the political left.”

In six of seven countries examined, mainstream right-wing political content received greater algorithmic amplification, while right-leaning news sources were also favoured in the U.S. media analysis.

Importantly, however, the study did not find evidence that algorithms systematically amplified far-right groups more than moderate groups.

That qualification is important.

It means we shouldn’t claim:

“The algorithm was designed to radicalize people into the far right.”

The evidence doesn’t support that.

The better claim is:

Engagement-based recommendation systems can create asymmetric amplification and provide highly efficient distribution mechanisms for political actors who are particularly effective at producing emotionally engaging content.

That is a much stronger argument.

---

XI. The transformation therefore looks like this

I would visualize our eventual argument as:

2014

Facebook behavioural data

↓

Psychometric profiling

↓

Personality → voter matching

↓

Microtargeting

↓

2015–16

Political campaigns learn to exploit individualized digital persuasion

↓

Alt-right develops sophisticated meme/network culture

↓

Russian IRA deploys computational propaganda

↓

Trump campaign exploits grievance + identity + social-media dynamics

↓

2016 election

↓

2017–2020

Political polarization becomes increasingly networked

↓

Algorithmic recommendation + engagement optimization

↓

Extremist / conspiratorial / populist narratives acquire enormous distribution

↓

2020–2026

The techniques no longer require Cambridge Analytica

↓

They become embedded in the political communication environment itself.

That last step is, in my view, the most important one.

---

XII. And there is evidence that the effect extends beyond the United States

This is where your earlier interest in Europe becomes particularly useful.

For example, research on Germany identifies PEGIDA, founded as a Facebook group in October 2014, as an early example of a major European right-wing populist movement whose organization and expansion were closely tied to social media.

Recent European research also finds that the far right can influence the mainstream political agenda, particularly around immigration, asylum and cultural identity. In Germany, the period from 2016–2018 saw especially intense public debate around these issues alongside increased attention to the AfD.

So we can make a much broader claim:

The significance of Cambridge Analytica lies not simply in what one company did during one American election, but in its place within the emergence of a new political communications environment that has subsequently been exploited by populist and far-right actors across democracies.

---

XIII. One crucial counterargument we should explicitly include

This will make the eventual case much stronger.

The Network Propaganda research from Oxford/Harvard is an excellent corrective to the simplistic Cambridge Analytica narrative.

Its authors examined four million political stories covering the period from April 2015 through January 2018.

Their conclusion was essentially that Cambridge Analytica’s claims about its own extraordinary influence were far less impressive than the company claimed.

This is actually useful to us.

Because our thesis doesn’t need:

“Cambridge Analytica secretly engineered Trump’s victory.”

It needs:

“Cambridge Analytica was an early, unusually visible expression of a technological transformation in political influence.”

The second proposition survives the first proposition being false.

---

The evidence hierarchy

I’d therefore rank the propositions we’re building as follows:

Proposition	Evidence
The operation began in 2014	Very strong
SCL contracted GSR in June 2014	Very strong
Facebook data were collected from users and their friends	Very strong
Psychometric models were constructed	Very strong
Models were connected to voter information	Very strong
The objective involved political targeting	Very strong
These methods can reasonably be characterized as psychological operations	Strong academic support
Cambridge Analytica used the system in the 2016 campaign	Strong
Data-driven microtargeting was part of the 2016 political environment	Very strong
Social media helped amplify divisive political content	Very strong
Russian actors exploited the same ecosystem	Very strong
Alt-right movements exploited networked social media	Very strong
Social-media mechanisms contributed to normalization/polarization	Strong
Cambridge Analytica caused the alt-right	Not established
Cambridge Analytica caused Trump’s victory	Not established
The technological paradigm pioneered/embodied by CA persists today	Strong
That paradigm contributes to today’s political information environment	Strong, but multifactorial

And I think that last distinction is exactly what will prevent the argument from becoming a conspiracy theory while still allowing us to make a very serious case about psychological influence, political manipulation and the transformation of public discourse.

The primary sources I would build the project around

• UK Parliament — Digital, Culture, Media and Sport Committee, Disinformation and ‘Fake News’: probably our single best source for the 2014 chronology and the actual SCL/GSR contract. UK Parliament report
• U.S. Federal Trade Commission — Kogan/Nix complaint: excellent legal documentation of the Facebook API, 2014 agreement and data acquisition. FTC complaint
• UK Information Commissioner’s Office — investigation into political data analytics: particularly valuable for connecting the harvested data to political targeting. ICO investigation
• Vian Bakir, Frontiers in Communication: our principal scholarly source for the “psychological operations” argument. Peer-reviewed study on Cambridge Analytica and psy-ops
• U.S. Senate Select Committee on Intelligence: essential for the Russian computational-propaganda component. Senate hearing on social-media influence
• Oxford/Harvard, Network Propaganda: essential counterweight preventing us from overstating Cambridge Analytica’s direct electoral effect. Network Propaganda research
• PNAS, Algorithmic amplification of politics on Twitter: excellent evidence for the post-2016 evolution of the information environment. PNAS study

The really interesting next step is to reconstruct 2014–2016 almost month by month. In particular, I’d like to trace SCL → GSR/Kogan → the 2014 psychometric experiment → the Mercer/Republican network → Cambridge Analytica’s emergence → Ted Cruz → Trump → Breitbart/alt-right → Russian IRA → election day, and identify exactly where we have documentary evidence versus inference. That would let us build a rigorous causal narrative rather than simply assembling quotations that happen to support the thesis.

OK, I concur. Please elaborate in part II of this.

copyblock.wordpress.com

2026-09-07 22:44:20

How the F-K did we get here? An Analysis and Timeline of the Rise of the Alt-Right Movement and the Role of Social Media, Part II

Source: Cambridge Analytica Evidence Sources – body of evidence from my personal observations and sources and corroborating evidence aggregated with the help of ChatGPT.

Part I of this analysis concerning the disturbing political and economic situation in Western countries, predominantly the US, but at an increasing rate and reach in Europe and other parts of the world as well, looks at the conditions that helped to create an interconnected network of powerful factions in the political and technological arena, who have been taking advantage of their leverage specific to their particular expertise. Part II now puts these significant and interconnected events and developments on a timeline.

yes, please elaborate.

Absolutely. I would now turn this into a chronological causal reconstruction, rather than treating “Cambridge Analytica” as a single event. The evidence becomes much more compelling when we show how several strands developed in parallel and then converged in 2016.

There is also one important refinement: the story begins before Cambridge Analytica existed as a U.S. political brand. The institutional lineage is SCL → SCL Elections → Cambridge Analytica, with the underlying data/psychometric work beginning in 2014. Christopher Wylie told the UK Parliament that Cambridge Analytica was essentially a U.S. vehicle/brand created around SCL’s existing work, with Robert Mercer becoming its majority shareholder and SCL licensing its intellectual property to it.

1. 2013–early 2014: the technological idea

Photo by Paul Hanaoka on Unsplash

The first component is psychological inference from digital behaviour.

The FTC’s subsequent investigation found that, by late 2013 or early 2014, Alexander Nix, SCL Elections and Cambridge Analytica had become interested in research showing that Facebook “likes” could be used to predict personality characteristics using the OCEAN model:

• Openness
• Conscientiousness
• Extraversion
• Agreeableness
• Neuroticism

The important point is that they weren’t interested in this primarily as academic psychology. The FTC says their objective was to commercialize it for “voter profiling, microtargeting, and other marketing services to U.S. campaigns.”

So the first conceptual transformation was:

ordinary Facebook behaviour

→ psychological inference

→ political intelligence

→ potentially individualized persuasion

That is the technological foundation.

---

2. Spring 2014: SCL turns the idea into an operational project

Aleksandr Kogan established Global Science Research (GSR) in spring 2014.

The UK Parliament’s investigation records that GSR developed an app designed to collect individual-level Facebook data. At the time Facebook’s “friends permissions” allowed an application to access information not only from the person who installed the app, but also from that person’s friends—without those friends necessarily knowing about it.

This is the crucial structural vulnerability.

It meant that:

one person’s participation could turn into access to a network of people who had never participated.

The FTC subsequently found that the GSR app ultimately collected data from roughly 250,000–270,000 U.S. app users and 50–65 million of their Facebook friends.

The later UK ICO investigation put the global figure at up to 87 million Facebook users, including approximately one million in Britain.

---

3. April–June 2014: the data architecture is built

This is perhaps the most important period in the entire chronology.

Facebook announced in April 2014 that it would stop allowing new applications to access friends’ data.

But existing applications were grandfathered for approximately another year.

The FTC explicitly says that this made Kogan’s existing application particularly attractive to Nix, SCL and Cambridge Analytica.

Then came the agreement between GSR and SCL Elections.

The FTC documents the sequence:

early 2014

SCL/Nix ↔ Kogan

↓

spring 2014

GSR established

↓

summer 2014

GSR app deployed

↓

Facebook data collected

↓

personality algorithm trained

↓

personality scores matched to voter records

↓

data used for voter profiling and targeted advertising.

This is not speculative. The FTC’s enforcement action established this as part of its case.

---

4. 2014: simultaneously, SCL is building the political operating system

This part is easy to overlook.

While the Kogan project was producing psychological data, SCL was also working with AggregateIQ (AIQ) on something called Ripon.

The UK Parliament describes Ripon as a political customer-relationship-management system developed for the U.S. market.

It was designed to help campaigns:

• identify voters;
• canvass them door-to-door;
• conduct telephone banking;
• record their opinions;
• manage campaign contacts.

The ICO independently confirms that SCL approached AIQ in early 2014 to develop the political CRM system for the U.S. 2014 midterms. Work began in April 2014.

This gives us something much more substantial than “Cambridge Analytica had Facebook data.”

There were two complementary systems being constructed:

System A — psychological intelligence

Who is this person?

System B — political operations

How do we act on that information?

Put those together and you have:

identify → classify → target → contact → measure → modify

That is the architecture of a modern influence operation.

---

5. Late 2014: Cambridge Analytica becomes the political vehicle

This is where Robert Mercer and Steve Bannon enter the story.

Christopher Wylie’s testimony to Parliament is unusually useful here. He described Cambridge Analytica as essentially a vehicle created around SCL’s intellectual property for the U.S. market, with Robert Mercer becoming the majority shareholder and SCL licensing its IP back to SCL Elections.

Contemporary reporting also documents the Mercer/Bannon/Nix relationship and the effort to bring SCL’s data-driven political methodology into U.S. conservative politics.

This is the first major political convergence:

SCL’s influence-operations expertise

psychometric Facebook research

political CRM technology

Mercer’s financial resources

Bannon’s political strategy

=

Cambridge Analytica

And this is where the story starts becoming relevant to the alt-right.

Not because Cambridge Analytica invented the alt-right—it did not—but because the same political network was simultaneously building a data operation and a media/political movement on the nationalist right.

---

6. 2014–2015: the alt-right/media component develops in parallel

This is the piece I think we should emphasize much more strongly than we did initially.

Cambridge Analytica wasn’t operating in an ideological vacuum.

Robert Mercer was also supporting Breitbart and Bannon’s political-media project.

The Oxford Network Propaganda research is particularly valuable here because it reconstructs the actual media network rather than relying on individual anecdotes.

Its analysis of millions of political stories and social-media interactions found that by 2015–16 the American right had developed a highly asymmetric media ecosystem, with Breitbart becoming extraordinarily central within the right-wing online network.

And something even more important happened:

Breitbart functioned as a bridge.

The researchers found that Breitbart helped transfer themes from the more extreme fringes of the right into the broader conservative ecosystem.

Their analysis of immigration coverage found that Breitbart connected:

white-nationalist narratives

↓

Breitbart

↓

mainstream conservative media

↓

Trump campaign discourse

That is a very important mechanism for our thesis.

The transformation isn’t:

“Cambridge Analytica created the alt-right.”

It is:

An increasingly sophisticated digital political infrastructure emerged in which data-driven targeting and an increasingly radicalized media ecosystem could operate together.

---

7. 2015: Ted Cruz becomes the first major U.S. test

This is an important bridge between the 2014 experiment and Trump.

By 2015, Cambridge Analytica was working for Ted Cruz’s presidential campaign, backed by the Mercer network.

Contemporary reporting documented that Cambridge Analytica was using the Facebook-derived data and psychometric approach to construct voter profiles for Cruz.

This is important because it demonstrates that:

the technology was not developed retrospectively to explain Trump’s victory.

It was already being deployed in an actual Republican presidential primary.

At this point, however, we should resist another tempting oversimplification.

The evidence does not establish that Cambridge Analytica’s psychographic model made Cruz competitive.

In fact, evidence later emerged that the company itself substantially overstated the effectiveness of its psychographic methods.

That matters.

---

8. Late 2015: the Facebook scandal becomes public

In December 2015, The Guardian revealed that SCL/Cambridge Analytica had acquired Facebook-derived data through Kogan.

This is significant because it means that the basic operation was publicly known before the 2016 election, although the full scale and significance weren’t understood at the time.

Senator Chuck Grassley’s later account of the Senate investigation explicitly notes:

“The Facebook story first broke in December 2015”

and identifies Cambridge Analytica’s subsequent work for Trump as the major development that followed.

So the chronology becomes:

2014

data operation begins

↓

2015

data/psychometric system used for Republican politics

↓

December 2015

public exposure

↓

2016

Trump campaign enters the picture.

---

9. Early 2016: the political ecosystem converges

Now three different things are happening simultaneously.

A. Cambridge Analytica

Data-driven political targeting.

B. Breitbart / Bannon / Mercer

A radicalizing right-wing media ecosystem.

C. Trump

A candidate whose political communication is extraordinarily well adapted to the emerging social-media environment.

The Oxford Network Propaganda research provides compelling evidence about the second and third elements.

During the 2016 primaries, Breitbart became increasingly central within the right-wing media ecosystem. By February 2016 it had eclipsed Fox News in the network of Twitter attention examined by the researchers.

And the content wasn’t simply “conservative.”

The researchers found that Breitbart was especially important in constructing narratives around:

• immigration;
• Muslims;
• terrorism;
• corruption;
• establishment betrayal;
• racial and cultural identity.

That is essentially the ideological substrate into which Trump’s campaign inserted itself.

---

10. Then comes the crucial June 2016 transition

Trump’s campaign hired Cambridge Analytica.

Brad Parscale later explained to PBS’s Frontline that Cambridge supplied staff to the campaign and that he was particularly interested in the data-science personnel.

Cambridge staff were embedded in the campaign’s San Antonio operation.

A contemporary investigation reported that the company eventually had around 13 people working with the campaign.

The Washington Post reported that Cambridge’s system supplied recommendations concerning:

• persuadable voters;
• donors;
• rally locations;
• volunteers;
• advertising locations.

This is where our psychological warfare argument becomes operational.

The campaign could theoretically move from:

“What does the electorate believe?”

to:

“Which people are persuadable?”

to:

“Which message should this particular type of person receive?”

That is a fundamentally different model of political persuasion.

---

11. But here’s the fascinating contradiction

By late September/early October 2016, the Trump campaign reportedly decided to rely primarily on Republican National Committee data, rather than Cambridge Analytica’s data.

CBS reported that the RNC data were considered substantially more accurate and became the main data source for the general election.

This is enormously important for intellectual honesty.

It means:

We cannot responsibly say:

“Cambridge Analytica’s Facebook profiles were the secret weapon that won Trump the presidency.”

The evidence doesn’t support that.

But something more subtle is supported:

Cambridge Analytica helped establish and operationalize a political model in which the electorate was treated as a collection of individually targetable psychological and behavioural segments.

And that model survived the failure or obsolescence of Cambridge’s particular data set.

---

12. Meanwhile, another psychological operation was already running

This is where the Russian Internet Research Agency becomes essential.

The bipartisan Senate Intelligence Committee concluded that Russia’s IRA operation was part of a broader information-warfare campaign designed to “sow discord” in American society.

Importantly, the committee found that the operation began before the election and wasn’t simply an election-day operation. It exploited existing divisions involving:

• race;
• immigration;
• gun rights;
• religion;
• identity.

The Senate’s detailed research also found that IRA activity increased substantially around major political events throughout 2015–17.

And the actual content strategy is extraordinarily revealing.

During election week 2016, the IRA targeted different constituencies with different psychological narratives:

right-wing audiences

→ anger
→ suspicion
→ conspiracy
→ voter fraud
→ fear that Clinton would “steal” the election

African-American audiences

→ alienation
→ police brutality
→ political disengagement
→ “don’t vote”

left-wing audiences

→ identity
→ anti-establishment sentiment
→ distrust of Clinton.

This is psychological segmentation.

And notice the conceptual similarity to Cambridge Analytica:

Cambridge Analytica

identify psychological characteristics → tailor political communication.

IRA

identify social identities → tailor divisive political communication.

The two operations were not coordinated, and we should absolutely not imply that they were.

But they were exploiting the same fundamental property of social media:

different populations can be identified and exposed to different realities.

---

13. This is where the concept of “informational dominance” becomes useful

A submission to the UK Parliament during the Cambridge Analytica investigation described information operations in terms of two components:

“informational dominance”

and the collection of information about targets in order to identify and exploit mental vulnerabilities to provoke behaviour.

I would be careful with this source because it is written evidence submitted to Parliament, not an official government finding.

But conceptually it captures what we’re trying to describe.

The important shift is:

Traditional propaganda

One message → many people

versus

Data-driven influence

Many people → classified into psychological/social groups → different messages → behavioural feedback

That is the fundamental technological transformation.

---

14. And then something even more important happens: the system becomes self-reinforcing

This is where we move from Cambridge Analytica to today.

The Oxford Network Propaganda research found a striking asymmetry in the American media system.

The right-wing ecosystem was:

• more insulated;
• more internally interconnected;
• more oriented toward extreme sources;
• heavily centered around Breitbart;
• increasingly capable of transmitting stories outward into mainstream discourse.

That creates a feedback loop:

fringe narrative

↓

Breitbart

↓

Trump

↓

mainstream media coverage

↓

social-media reaction

↓

more attention to Breitbart

↓

more extreme content

↓

mainstream response

↓

more attention

This is not necessarily persuasion in the traditional sense.

It is agenda formation.

---

15. That may actually be more important than microtargeting

This is the point I would make the centerpiece of the eventual essay.

Cambridge Analytica’s most important historical contribution may not have been:

“We found the psychological button that makes people vote Trump.”

It may have been:

“Political communication can now be organized around the continuous collection, classification and exploitation of behavioural data.”

And the political ecosystem subsequently discovered something even more powerful:

You don’t necessarily have to persuade people individually if you can alter the information environment in which everyone makes political judgments.

That is where:

Cambridge Analytica

meets

Breitbart

meets

Trump

meets

Russian information warfare

meets

social-media algorithms

meets

today’s political discourse.

---

16. The post-election phase is crucial

One of the strongest pieces of evidence against treating this as merely a 2016 election story is the Senate Intelligence Committee’s finding that IRA activity actually increased after Election Day.

The committee found significant post-election increases in activity:

• Instagram: +238%
• Facebook: +59%
• Twitter: +52%
• YouTube: +84%

That tells us something profound.

The objective wasn’t simply:

Get Trump elected.

The information operation continued because the underlying objective—political polarization and social discord—continued to have value.

That is exactly the bridge from election interference to long-term political discourse.

---

17. The resulting causal model

I would now formulate our overall model like this:

PHASE I — TECHNOLOGICAL DEVELOPMENT

Photo by Joshua Sortino on Unsplash

2013–2014

Facebook behavioural data

→ psychometrics

→ personality inference

→ voter identification

→ political CRM

PHASE II — POLITICAL INDUSTRIALIZATION

2014–2015

SCL + Mercer + Bannon

→ Cambridge Analytica

→ Republican campaigns

→ Cruz

→ political microtargeting

PHASE III — IDEOLOGICAL CONVERGENCE

2015–2016

Breitbart

→ Bannon

→ Mercer

→ immigration / Islam / nationalism / anti-establishment narratives

→ alt-right enters mainstream political discourse

PHASE IV — MASSIVE CONVERGENCE

2016

Trump

Cambridge Analytica

Breitbart

Facebook

Russian IRA

algorithmic distribution

↓

2016 election

PHASE V — NORMALIZATION

2017–present

The technology no longer belongs to Cambridge Analytica.

The techniques become common political infrastructure.

The political vocabulary becomes increasingly polarized.

The networks persist.

The audiences remain segmented.

The platforms retain behavioural data.

And political actors learn from what worked.

---

18. But there is an important distinction between radicalization and amplification

I would make this distinction explicit in our eventual paper.

There are at least three mechanisms:

① Persuasion

“I didn’t believe X. Now I believe X.”

Evidence for Cambridge Analytica’s effectiveness here is weak/moderate.

② Mobilization

“I already believe X. Now I’m politically active.”

Evidence here is substantially stronger.

③ Normalization

“I don’t necessarily believe X, but I now think lots of people believe X—and therefore X is legitimate to discuss.”

Evidence for this mechanism in the contemporary extreme-right information environment is particularly interesting.

And normalization can produce a fourth effect:

④ Agenda displacement

“The political system now has to talk about X.”

This is where the Breitbart → Trump → mainstream media relationship becomes extraordinarily important.

The Oxford research provides strong evidence for precisely this kind of agenda-setting dynamic.

---

19. So what exactly can we claim about the rise of the alt-right?

I’d now formulate the proposition very carefully:

Strong claim

Cambridge Analytica was part of the same political-financial-media network that helped bring data-driven political influence techniques into the emerging Trump-era conservative movement.

Supported.

Strong claim

The 2014–16 SCL/Cambridge Analytica project pioneered an operational model combining behavioural data, psychological profiling, voter records and targeted political communication.

Strongly supported.

Strong claim

The Trump-era right simultaneously developed a highly networked media ecosystem centered on Breitbart that provided a bridge between more extreme ideological positions and mainstream conservative politics.

Strongly supported by the Network Propaganda research.

Strong claim

Russian information operations exploited the same social-media environment through highly differentiated psychological and identity-based targeting.

Very strongly supported by the Senate investigation.

Reasonable inference

These developments mutually reinforced a political communication environment unusually favorable to populist, nationalist and far-right narratives.

Strong historical inference.

Claim we should not make

“Cambridge Analytica caused the alt-right.”

Not established.

Claim we probably can make

“Cambridge Analytica was an important early component of the technological and political infrastructure through which the alt-right and adjacent populist movements could scale, target, normalize and disseminate their ideas.”

That’s the proposition I’d defend.

---

20. And there’s one final twist that I think makes the story considerably more powerful

Cambridge Analytica disappeared in 2018. The phenomenon it represented did not.

The company went bankrupt.

But:

• political microtargeting remained;
• behavioural advertising remained;
• voter databases remained;
• social-media profiling remained;
• algorithmic recommendation remained;
• influencer networks remained;
• Breitbart remained;
• alternative media remained;
• computational propaganda remained;
• the political incentives to exploit outrage remained.

The Senate found that Russian social-media activity actually increased after the 2016 election.

So the historical argument should perhaps end not with:

“Cambridge Analytica manipulated the 2016 election.”

but with something considerably more consequential:

2016 was the moment when democratic politics discovered that the electorate could be treated simultaneously as a population to be measured, a collection of psychological profiles to be targeted, and a network of identities to be manipulated. Cambridge Analytica was one of the earliest and most visible manifestations of that transformation. The infrastructure survived the company itself.

That is, I think, the real story.

---

The next stage: build the documentary timeline

I suggest we now go one level deeper and construct a 2014–2016 evidence table with individual events, something like:

Date	Event	Actors	What actually happened	Evidence	What it proves
Jan–Mar 2014	SCL–Kogan contact	Nix, SCL, Kogan	Psychometric collaboration discussed	FTC / Parliament	Origin
Apr 2014	Facebook API change	Facebook	Friends’ data access restricted but grandfathered	FTC	Enabling vulnerability
Apr 2014	Ripon begins	SCL/AIQ	Political CRM developed	ICO	Operational infrastructure
Jun 2014	GSR/SCL agreement	Kogan/SCL	Facebook + psychometrics + voter data	FTC	Core experiment
Summer 2014	Data collection	GSR/SCL	Millions of profiles generated	FTC	Scale
Late 2014	Mercer/Bannon involvement	Mercer/Bannon/Nix	CA political vehicle established	Parliament	Political network
2015	Cruz campaign	CA/Cruz	Psychographic political targeting	contemporary reporting	Deployment
Dec 2015	Guardian exposes data	Guardian/SCL	Facebook data story becomes public	Senate record	Public disclosure
2015–16	Breitbart expansion	Bannon/Mercer	Right-wing network becomes increasingly dominant	Oxford	Ideological infrastructure
2016	Trump hires CA	Trump/Parscale/CA	CA embeds in campaign	PBS/WaPo	Operational deployment
2016	Russian IRA	IRA	Segmented influence operation	Senate	Parallel psy-op
Nov. 2016	Election	—	Trump wins	—	Convergence point
2017–18	Network persists	IRA/right-wing media	Political influence continues	Senate/Oxford	Long-term effect

That table would let us do something particularly valuable: assign an evidentiary grade to every link in the causal chain—documented fact / strong inference / contested / unsupported.

That, in turn, would make the eventual argument much harder to dismiss as either partisan rhetoric or a conspiracy theory..

Thank you, ChatGPT. I rest my case in the way you suggest. That’s a good enough position I can concur with.

Photo by Paul Hanaoka on Unsplash

This hallway in the Reina Sofia reminded me of the Death Star. – Download this photo by Paul Hanaoka on Unsplash

^unsplash.com

Source: Cambridge Analytica Evidence Sources – body of evidence from my personal observations and sources and corroborating evidence aggregated with the help of ChatGPT.

Part I of this analysis concerning the disturbing political and economic situation in Western countries, predominantly the US, but at an increasing rate and reach in Europe and other parts of the world as well, looks at the conditions that helped to create an interconnected network of powerful factions in the political and technological arena, who have been taking advantage of their leverage specific to their particular expertise. Part II now puts these significant and interconnected events and developments on a timeline.

yes, please elaborate.

Absolutely. I would now turn this into a chronological causal reconstruction, rather than treating “Cambridge Analytica” as a single event. The evidence becomes much more compelling when we show how several strands developed in parallel and then converged in 2016.

There is also one important refinement: the story begins before Cambridge Analytica existed as a U.S. political brand. The institutional lineage is SCL → SCL Elections → Cambridge Analytica, with the underlying data/psychometric work beginning in 2014. Christopher Wylie told the UK Parliament that Cambridge Analytica was essentially a U.S. vehicle/brand created around SCL’s existing work, with Robert Mercer becoming its majority shareholder and SCL licensing its intellectual property to it.

1. 2013–early 2014: the technological idea

Photo by Paul Hanaoka on Unsplash

The first component is psychological inference from digital behaviour.

The FTC’s subsequent investigation found that, by late 2013 or early 2014, Alexander Nix, SCL Elections and Cambridge Analytica had become interested in research showing that Facebook “likes” could be used to predict personality characteristics using the OCEAN model:

• Openness
• Conscientiousness
• Extraversion
• Agreeableness
• Neuroticism

The important point is that they weren’t interested in this primarily as academic psychology. The FTC says their objective was to commercialize it for “voter profiling, microtargeting, and other marketing services to U.S. campaigns.”

So the first conceptual transformation was:

ordinary Facebook behaviour

→ psychological inference

→ political intelligence

→ potentially individualized persuasion

That is the technological foundation.

---

2. Spring 2014: SCL turns the idea into an operational project

Aleksandr Kogan established Global Science Research (GSR) in spring 2014.

The UK Parliament’s investigation records that GSR developed an app designed to collect individual-level Facebook data. At the time Facebook’s “friends permissions” allowed an application to access information not only from the person who installed the app, but also from that person’s friends—without those friends necessarily knowing about it.

This is the crucial structural vulnerability.

It meant that:

one person’s participation could turn into access to a network of people who had never participated.

The FTC subsequently found that the GSR app ultimately collected data from roughly 250,000–270,000 U.S. app users and 50–65 million of their Facebook friends.

The later UK ICO investigation put the global figure at up to 87 million Facebook users, including approximately one million in Britain.

---

3. April–June 2014: the data architecture is built

This is perhaps the most important period in the entire chronology.

Facebook announced in April 2014 that it would stop allowing new applications to access friends’ data.

But existing applications were grandfathered for approximately another year.

The FTC explicitly says that this made Kogan’s existing application particularly attractive to Nix, SCL and Cambridge Analytica.

Then came the agreement between GSR and SCL Elections.

The FTC documents the sequence:

early 2014

SCL/Nix ↔ Kogan

↓

spring 2014

GSR established

↓

summer 2014

GSR app deployed

↓

Facebook data collected

↓

personality algorithm trained

↓

personality scores matched to voter records

↓

data used for voter profiling and targeted advertising.

This is not speculative. The FTC’s enforcement action established this as part of its case.

---

4. 2014: simultaneously, SCL is building the political operating system

This part is easy to overlook.

While the Kogan project was producing psychological data, SCL was also working with AggregateIQ (AIQ) on something called Ripon.

The UK Parliament describes Ripon as a political customer-relationship-management system developed for the U.S. market.

It was designed to help campaigns:

• identify voters;
• canvass them door-to-door;
• conduct telephone banking;
• record their opinions;
• manage campaign contacts.

The ICO independently confirms that SCL approached AIQ in early 2014 to develop the political CRM system for the U.S. 2014 midterms. Work began in April 2014.

This gives us something much more substantial than “Cambridge Analytica had Facebook data.”

There were two complementary systems being constructed:

System A — psychological intelligence

Who is this person?

System B — political operations

How do we act on that information?

Put those together and you have:

identify → classify → target → contact → measure → modify

That is the architecture of a modern influence operation.

---

5. Late 2014: Cambridge Analytica becomes the political vehicle

This is where Robert Mercer and Steve Bannon enter the story.

Christopher Wylie’s testimony to Parliament is unusually useful here. He described Cambridge Analytica as essentially a vehicle created around SCL’s intellectual property for the U.S. market, with Robert Mercer becoming the majority shareholder and SCL licensing its IP back to SCL Elections.

Contemporary reporting also documents the Mercer/Bannon/Nix relationship and the effort to bring SCL’s data-driven political methodology into U.S. conservative politics.

This is the first major political convergence:

SCL’s influence-operations expertise

psychometric Facebook research

political CRM technology

Mercer’s financial resources

Bannon’s political strategy

=

Cambridge Analytica

And this is where the story starts becoming relevant to the alt-right.

Not because Cambridge Analytica invented the alt-right—it did not—but because the same political network was simultaneously building a data operation and a media/political movement on the nationalist right.

---

6. 2014–2015: the alt-right/media component develops in parallel

This is the piece I think we should emphasize much more strongly than we did initially.

Cambridge Analytica wasn’t operating in an ideological vacuum.

Robert Mercer was also supporting Breitbart and Bannon’s political-media project.

The Oxford Network Propaganda research is particularly valuable here because it reconstructs the actual media network rather than relying on individual anecdotes.

Its analysis of millions of political stories and social-media interactions found that by 2015–16 the American right had developed a highly asymmetric media ecosystem, with Breitbart becoming extraordinarily central within the right-wing online network.

And something even more important happened:

Breitbart functioned as a bridge.

The researchers found that Breitbart helped transfer themes from the more extreme fringes of the right into the broader conservative ecosystem.

Their analysis of immigration coverage found that Breitbart connected:

white-nationalist narratives

↓

Breitbart

↓

mainstream conservative media

↓

Trump campaign discourse

That is a very important mechanism for our thesis.

The transformation isn’t:

“Cambridge Analytica created the alt-right.”

It is:

An increasingly sophisticated digital political infrastructure emerged in which data-driven targeting and an increasingly radicalized media ecosystem could operate together.

---

7. 2015: Ted Cruz becomes the first major U.S. test

This is an important bridge between the 2014 experiment and Trump.

By 2015, Cambridge Analytica was working for Ted Cruz’s presidential campaign, backed by the Mercer network.

Contemporary reporting documented that Cambridge Analytica was using the Facebook-derived data and psychometric approach to construct voter profiles for Cruz.

This is important because it demonstrates that:

the technology was not developed retrospectively to explain Trump’s victory.

It was already being deployed in an actual Republican presidential primary.

At this point, however, we should resist another tempting oversimplification.

The evidence does not establish that Cambridge Analytica’s psychographic model made Cruz competitive.

In fact, evidence later emerged that the company itself substantially overstated the effectiveness of its psychographic methods.

That matters.

---

8. Late 2015: the Facebook scandal becomes public

In December 2015, The Guardian revealed that SCL/Cambridge Analytica had acquired Facebook-derived data through Kogan.

This is significant because it means that the basic operation was publicly known before the 2016 election, although the full scale and significance weren’t understood at the time.

Senator Chuck Grassley’s later account of the Senate investigation explicitly notes:

“The Facebook story first broke in December 2015”

and identifies Cambridge Analytica’s subsequent work for Trump as the major development that followed.

So the chronology becomes:

2014

data operation begins

↓

2015

data/psychometric system used for Republican politics

↓

December 2015

public exposure

↓

2016

Trump campaign enters the picture.

---

9. Early 2016: the political ecosystem converges

Now three different things are happening simultaneously.

A. Cambridge Analytica

Data-driven political targeting.

B. Breitbart / Bannon / Mercer

A radicalizing right-wing media ecosystem.

C. Trump

A candidate whose political communication is extraordinarily well adapted to the emerging social-media environment.

The Oxford Network Propaganda research provides compelling evidence about the second and third elements.

During the 2016 primaries, Breitbart became increasingly central within the right-wing media ecosystem. By February 2016 it had eclipsed Fox News in the network of Twitter attention examined by the researchers.

And the content wasn’t simply “conservative.”

The researchers found that Breitbart was especially important in constructing narratives around:

• immigration;
• Muslims;
• terrorism;
• corruption;
• establishment betrayal;
• racial and cultural identity.

That is essentially the ideological substrate into which Trump’s campaign inserted itself.

---

10. Then comes the crucial June 2016 transition

Trump’s campaign hired Cambridge Analytica.

Brad Parscale later explained to PBS’s Frontline that Cambridge supplied staff to the campaign and that he was particularly interested in the data-science personnel.

Cambridge staff were embedded in the campaign’s San Antonio operation.

A contemporary investigation reported that the company eventually had around 13 people working with the campaign.

The Washington Post reported that Cambridge’s system supplied recommendations concerning:

• persuadable voters;
• donors;
• rally locations;
• volunteers;
• advertising locations.

This is where our psychological warfare argument becomes operational.

The campaign could theoretically move from:

“What does the electorate believe?”

to:

“Which people are persuadable?”

to:

“Which message should this particular type of person receive?”

That is a fundamentally different model of political persuasion.

---

11. But here’s the fascinating contradiction

By late September/early October 2016, the Trump campaign reportedly decided to rely primarily on Republican National Committee data, rather than Cambridge Analytica’s data.

CBS reported that the RNC data were considered substantially more accurate and became the main data source for the general election.

This is enormously important for intellectual honesty.

It means:

We cannot responsibly say:

“Cambridge Analytica’s Facebook profiles were the secret weapon that won Trump the presidency.”

The evidence doesn’t support that.

But something more subtle is supported:

Cambridge Analytica helped establish and operationalize a political model in which the electorate was treated as a collection of individually targetable psychological and behavioural segments.

And that model survived the failure or obsolescence of Cambridge’s particular data set.

---

12. Meanwhile, another psychological operation was already running

This is where the Russian Internet Research Agency becomes essential.

The bipartisan Senate Intelligence Committee concluded that Russia’s IRA operation was part of a broader information-warfare campaign designed to “sow discord” in American society.

Importantly, the committee found that the operation began before the election and wasn’t simply an election-day operation. It exploited existing divisions involving:

• race;
• immigration;
• gun rights;
• religion;
• identity.

The Senate’s detailed research also found that IRA activity increased substantially around major political events throughout 2015–17.

And the actual content strategy is extraordinarily revealing.

During election week 2016, the IRA targeted different constituencies with different psychological narratives:

right-wing audiences

→ anger
→ suspicion
→ conspiracy
→ voter fraud
→ fear that Clinton would “steal” the election

African-American audiences

→ alienation
→ police brutality
→ political disengagement
→ “don’t vote”

left-wing audiences

→ identity
→ anti-establishment sentiment
→ distrust of Clinton.

This is psychological segmentation.

And notice the conceptual similarity to Cambridge Analytica:

Cambridge Analytica

identify psychological characteristics → tailor political communication.

IRA

identify social identities → tailor divisive political communication.

The two operations were not coordinated, and we should absolutely not imply that they were.

But they were exploiting the same fundamental property of social media:

different populations can be identified and exposed to different realities.

---

13. This is where the concept of “informational dominance” becomes useful

A submission to the UK Parliament during the Cambridge Analytica investigation described information operations in terms of two components:

“informational dominance”

and the collection of information about targets in order to identify and exploit mental vulnerabilities to provoke behaviour.

I would be careful with this source because it is written evidence submitted to Parliament, not an official government finding.

But conceptually it captures what we’re trying to describe.

The important shift is:

Traditional propaganda

One message → many people

versus

Data-driven influence

Many people → classified into psychological/social groups → different messages → behavioural feedback

That is the fundamental technological transformation.

---

14. And then something even more important happens: the system becomes self-reinforcing

This is where we move from Cambridge Analytica to today.

The Oxford Network Propaganda research found a striking asymmetry in the American media system.

The right-wing ecosystem was:

• more insulated;
• more internally interconnected;
• more oriented toward extreme sources;
• heavily centered around Breitbart;
• increasingly capable of transmitting stories outward into mainstream discourse.

That creates a feedback loop:

fringe narrative

↓

Breitbart

↓

Trump

↓

mainstream media coverage

↓

social-media reaction

↓

more attention to Breitbart

↓

more extreme content

↓

mainstream response

↓

more attention

This is not necessarily persuasion in the traditional sense.

It is agenda formation.

---

15. That may actually be more important than microtargeting

This is the point I would make the centerpiece of the eventual essay.

Cambridge Analytica’s most important historical contribution may not have been:

“We found the psychological button that makes people vote Trump.”

It may have been:

“Political communication can now be organized around the continuous collection, classification and exploitation of behavioural data.”

And the political ecosystem subsequently discovered something even more powerful:

You don’t necessarily have to persuade people individually if you can alter the information environment in which everyone makes political judgments.

That is where:

Cambridge Analytica

meets

Breitbart

meets

Trump

meets

Russian information warfare

meets

social-media algorithms

meets

today’s political discourse.

---

16. The post-election phase is crucial

One of the strongest pieces of evidence against treating this as merely a 2016 election story is the Senate Intelligence Committee’s finding that IRA activity actually increased after Election Day.

The committee found significant post-election increases in activity:

• Instagram: +238%
• Facebook: +59%
• Twitter: +52%
• YouTube: +84%

That tells us something profound.

The objective wasn’t simply:

Get Trump elected.

The information operation continued because the underlying objective—political polarization and social discord—continued to have value.

That is exactly the bridge from election interference to long-term political discourse.

---

17. The resulting causal model

I would now formulate our overall model like this:

PHASE I — TECHNOLOGICAL DEVELOPMENT

Photo by Joshua Sortino on Unsplash

2013–2014

Facebook behavioural data

→ psychometrics

→ personality inference

→ voter identification

→ political CRM

PHASE II — POLITICAL INDUSTRIALIZATION

2014–2015

SCL + Mercer + Bannon

→ Cambridge Analytica

→ Republican campaigns

→ Cruz

→ political microtargeting

PHASE III — IDEOLOGICAL CONVERGENCE

2015–2016

Breitbart

→ Bannon

→ Mercer

→ immigration / Islam / nationalism / anti-establishment narratives

→ alt-right enters mainstream political discourse

PHASE IV — MASSIVE CONVERGENCE

2016

Trump

Cambridge Analytica

Breitbart

Facebook

Russian IRA

algorithmic distribution

↓

2016 election

PHASE V — NORMALIZATION

2017–present

The technology no longer belongs to Cambridge Analytica.

The techniques become common political infrastructure.

The political vocabulary becomes increasingly polarized.

The networks persist.

The audiences remain segmented.

The platforms retain behavioural data.

And political actors learn from what worked.

---

18. But there is an important distinction between radicalization and amplification

I would make this distinction explicit in our eventual paper.

There are at least three mechanisms:

① Persuasion

“I didn’t believe X. Now I believe X.”

Evidence for Cambridge Analytica’s effectiveness here is weak/moderate.

② Mobilization

“I already believe X. Now I’m politically active.”

Evidence here is substantially stronger.

③ Normalization

“I don’t necessarily believe X, but I now think lots of people believe X—and therefore X is legitimate to discuss.”

Evidence for this mechanism in the contemporary extreme-right information environment is particularly interesting.

And normalization can produce a fourth effect:

④ Agenda displacement

“The political system now has to talk about X.”

This is where the Breitbart → Trump → mainstream media relationship becomes extraordinarily important.

The Oxford research provides strong evidence for precisely this kind of agenda-setting dynamic.

---

19. So what exactly can we claim about the rise of the alt-right?

I’d now formulate the proposition very carefully:

Strong claim

Cambridge Analytica was part of the same political-financial-media network that helped bring data-driven political influence techniques into the emerging Trump-era conservative movement.

Supported.

Strong claim

The 2014–16 SCL/Cambridge Analytica project pioneered an operational model combining behavioural data, psychological profiling, voter records and targeted political communication.

Strongly supported.

Strong claim

The Trump-era right simultaneously developed a highly networked media ecosystem centered on Breitbart that provided a bridge between more extreme ideological positions and mainstream conservative politics.

Strongly supported by the Network Propaganda research.

Strong claim

Russian information operations exploited the same social-media environment through highly differentiated psychological and identity-based targeting.

Very strongly supported by the Senate investigation.

Reasonable inference

These developments mutually reinforced a political communication environment unusually favorable to populist, nationalist and far-right narratives.

Strong historical inference.

Claim we should not make

“Cambridge Analytica caused the alt-right.”

Not established.

Claim we probably can make

“Cambridge Analytica was an important early component of the technological and political infrastructure through which the alt-right and adjacent populist movements could scale, target, normalize and disseminate their ideas.”

That’s the proposition I’d defend.

---

20. And there’s one final twist that I think makes the story considerably more powerful

Cambridge Analytica disappeared in 2018. The phenomenon it represented did not.

The company went bankrupt.

But:

• political microtargeting remained;
• behavioural advertising remained;
• voter databases remained;
• social-media profiling remained;
• algorithmic recommendation remained;
• influencer networks remained;
• Breitbart remained;
• alternative media remained;
• computational propaganda remained;
• the political incentives to exploit outrage remained.

The Senate found that Russian social-media activity actually increased after the 2016 election.

So the historical argument should perhaps end not with:

“Cambridge Analytica manipulated the 2016 election.”

but with something considerably more consequential:

2016 was the moment when democratic politics discovered that the electorate could be treated simultaneously as a population to be measured, a collection of psychological profiles to be targeted, and a network of identities to be manipulated. Cambridge Analytica was one of the earliest and most visible manifestations of that transformation. The infrastructure survived the company itself.

That is, I think, the real story.

---

The next stage: build the documentary timeline

I suggest we now go one level deeper and construct a 2014–2016 evidence table with individual events, something like:

Date	Event	Actors	What actually happened	Evidence	What it proves
Jan–Mar 2014	SCL–Kogan contact	Nix, SCL, Kogan	Psychometric collaboration discussed	FTC / Parliament	Origin
Apr 2014	Facebook API change	Facebook	Friends’ data access restricted but grandfathered	FTC	Enabling vulnerability
Apr 2014	Ripon begins	SCL/AIQ	Political CRM developed	ICO	Operational infrastructure
Jun 2014	GSR/SCL agreement	Kogan/SCL	Facebook + psychometrics + voter data	FTC	Core experiment
Summer 2014	Data collection	GSR/SCL	Millions of profiles generated	FTC	Scale
Late 2014	Mercer/Bannon involvement	Mercer/Bannon/Nix	CA political vehicle established	Parliament	Political network
2015	Cruz campaign	CA/Cruz	Psychographic political targeting	contemporary reporting	Deployment
Dec 2015	Guardian exposes data	Guardian/SCL	Facebook data story becomes public	Senate record	Public disclosure
2015–16	Breitbart expansion	Bannon/Mercer	Right-wing network becomes increasingly dominant	Oxford	Ideological infrastructure
2016	Trump hires CA	Trump/Parscale/CA	CA embeds in campaign	PBS/WaPo	Operational deployment
2016	Russian IRA	IRA	Segmented influence operation	Senate	Parallel psy-op
Nov. 2016	Election	—	Trump wins	—	Convergence point
2017–18	Network persists	IRA/right-wing media	Political influence continues	Senate/Oxford	Long-term effect

That table would let us do something particularly valuable: assign an evidentiary grade to every link in the causal chain—documented fact / strong inference / contested / unsupported.

That, in turn, would make the eventual argument much harder to dismiss as either partisan rhetoric or a conspiracy theory..

Thank you, ChatGPT. I rest my case in the way you suggest. That’s a good enough position I can concur with.