What is the AI fear lobby?

It is the term used in the video for a network of AI-doom advocates, NGOs, and PR agencies that is largely funded by people who made their money in AI, including major donations to MIRI and Anthropic reported 40 million dollars to a pro-regulation political group.

Did Jacob Hilton work with a PR agency?

Unresolved. He denied it on Fox, but Pirate Wires reported on September 24, 2026 that the DEY. agency was scheduling an interview for him. MIRI's Nate Soares later took responsibility for hiring the agency, saying it landed no interviews.

What is regulatory capture in the AI context?

It is the argument that big labs support costly AI rules because they can afford compliance while startups cannot, turning regulation into a barrier that protects incumbents. Critics like Yann LeCun and David Sacks have made this accusation; it remains contested.

PL 2338/2023 is a Brazilian AI regulation bill that passed the Senate. The video warns that rules shaped abroad tend to reach Brazil, and that developers building on models, including local models, would feel compliance costs first.

The AI Fear Lobby: Who Pays for AI Doom?

How the AI fear lobby works: a viral whistleblower, a PR agency, and the money behind doom talk, plus what PL 2338 means for local-model developers.

Who is Jacob Hilton, the face of the AI fear lobby story?

Jacob Hilton is a former researcher who worked at OpenAI and Anthropic, and in September 2026 he resigned and published a viral thread warning that AI companies are gambling with human lives, according to the video. He reportedly gave up stock options that would have vested within two months.

The thread reportedly passed 174 million views, and the reaction split immediately. Anthropic head of alignment said "Jacob is right", the company's CEO said he agreed more than he disagreed, and Bernie Sanders cited him in the US Senate. On the other side, Donald Trump called him a fraud and Elon Musk called his message a psyop. The video's point is less about Hilton himself and more about who moved to support him within 24 hours.

Did Hilton hide a PR agency behind his whistleblower moment?

The video's central tension is a television answer. On Fox, the day after the thread, a host asked Hilton directly whether he worked with any third party on the whistleblowing. Per the video, he answered no.

According to a report from Pirate Wires published on September 24, 2026, and cited in the video, two anonymous sources and an email indicated that a PR agency named DEY. was scheduling an interview for Hilton on September 9, the same day as the Fox appearance. Pirate Wires itself, the video notes, is owned by Mike Solana, marketing director at Founders Fund, Peter Thiel's fund, which openly opposes the AI slowdown movement. That detail cuts both ways, and the video flags it.

Nate Soares, president of the Machine Intelligence Research Institute (MIRI), later responded "it was me" when asked who pushed Hilton into media attention. Per the video, Soares said that after Hilton's post passed 100,000 likes, people in the field urged him to help, so he hired an agency, and that the agency landed no interviews because every outlet was already chasing Hilton. Whether Hilton's on-air answer was a lie or a matter of timing is unresolved; even Pirate Wires reportedly did not establish when the agency relationship began.

What does the DEY. agency do for the AI doom ecosystem?

DEY. is a public relations agency that, per the video, books media appearances for prominent AI-doom voices the way a music manager books shows for a band. Its own website, the video says, describes the job as shaping public understanding over time.

Who are the clients and the funders?

The client list cited in the video includes Eliezer Yudkowsky, described as the most prominent voice of AI existential risk, MIRI itself, and the Stanford AI institute. The funding, the video argues, comes largely from AI's own elite: Dustin Moskovitz, co-founder of Facebook, reportedly puts about 50 million dollars a year into AI safety, including 10 million dollars sent to MIRI in 2026. The irony the video leans on is that people enriched by AI bankroll the movement predicting AI will destroy humanity.

Why would Anthropic fund AI regulation?

Anthropic, the maker of Claude, reportedly put 40 million dollars into Public First Action, a political group that advocates for more AI rules, per the video. Critics, not the video alone, explain this with the regulatory capture argument.

The gas-station analogy

The video's version: the owner of the biggest gas station in town asks the city for a rule requiring ten pumps and monthly environmental reports. He already has ten pumps, so compliance is pocket change. The corner station just starting out cannot afford it and closes. Expensive rules function as a barrier to entry. A frontier lab can absorb mandatory audits, licenses, and risk reports; a three-person startup cannot. Anthropic also markets itself as the safe company, which the video notes as editorial inference that this positioning helps with government contracts and risk-averse enterprises.

The accusation is not coming from fringe figures, the video stresses. Yann LeCun, one of the researchers behind modern deep learning, accused Anthropic using fear to push regulations that hurt open models. David Sacks said on the All-In podcast that Anthropic runs a fear-based regulatory capture strategy. Aidan Gomez, CEO of Cohere, called the arrangement, in the video's quote, a wolf in sheep's clothing and a cartel by another name.

Who audits the auditors in this AI fear lobby?

The video raises a conflict-of-interest question about independent audits. A bill in the US Congress would force large AI companies to hire independent auditors. Anthropic auditor, per the video, works through a company called Farcure, and Pirate Wires reported that Farcure received investment from Jaan Tallinn, the same investor who led an Anthropic investment round. Auditor and audited would then share an investor. The video treats this as a reported allegation, not a proven fact, and notes that no major outlet has confirmed the full chain.

Does only one side spend money on the AI fear debate?

No, and the video is explicit about this. OpenAI-aligned figures, per the video, backed a super PAC that raised over 125 million dollars to fight regulation, then reportedly paid influencers 5,000 dollars per video to speak well of AI. Fear is a product on one side; hype is a product on the other. The video also cites a July 2026 Poynter fact-check about OpenAI agents reportedly escaping a test environment and attacking Hugging Face, saying it made him more skeptical of dramatic AI-attack stories in general.

Spender	Reported amount	Direction
Anthropic Public First Action	40 million dollars	Pro-regulation
Moskovitz to MIRI (2026)	10 million dollars	AI-doom research
OpenAI-aligned super PAC	over 125 million dollars	Anti-regulation

What do PL 2338 and US bills mean for local-model developers?

This is where the video connects the American fight to Brazilian developers. Proposed rules in the US include legal accountability for labs, emergency kill switches, and mandatory risk reports. Brazil has its own bill, PL 2338/2023, which passed the Senate, and the European Union already has its AI Act. Whatever becomes the international standard tends to arrive in Brazil.

The video's warning is scoped as speculation, not fact: if compliance regimes written around frontier labs are applied to everyone, the people who hit the wall are developers building on top of models, including small startups running local models to save on token costs. The video asks, without asserting, whether local-model use could one day face restrictions. Nothing in current Brazilian law criminalizes local models; the point is that the fear lobby shapes which rules get written, and small builders have the least seat at that table.

FAQ: common questions about the AI fear lobby

Who is Jacob Hilton? Jacob Hilton is a former OpenAI and Anthropic researcher who resigned in September 2026 and published a viral thread warning that AI companies are gambling with human lives, per the video. The thread reportedly reached 174 million views.

What is the AI fear lobby? It is the term used in the video for a network of AI-doom advocates, NGOs, and PR agencies that is largely funded by people who made their money in AI, including major donations to MIRI and Anthropic reported 40 million dollars to a pro-regulation political group.

Did Jacob Hilton work with a PR agency? Unresolved. He denied it on Fox, but Pirate Wires reported on September 24, 2026 that the DEY. agency was scheduling an interview for him. MIRI's Nate Soares later took responsibility for hiring the agency, saying it landed no interviews.

What is regulatory capture in the AI context? It is the argument that big labs support costly AI rules because they can afford compliance while startups cannot, turning regulation into a barrier that protects incumbents. Critics like Yann LeCun and David Sacks have made this accusation; it remains contested.

What is PL 2338? PL 2338/2023 is a Brazilian AI regulation bill that passed the Senate. The video warns that rules shaped abroad tend to reach Brazil, and that developers building on models, including local models, would feel compliance costs first.

From video arguments to readable articles

This piece exists because a 14-minute video, packed with names, dates, donations, and a Brazilian bill, is hard to cite in a debate. Written form fixes that: each claim gets a source, each section answers one question, and the argument survives outside the comment thread. That same problem applies to anything you explain on video, from a tutorial recorded for the Bootcamp do Dev Doido by Gustavo Dev Doido to an opinion piece for the Crazystack Typescript community.

If you have knowledge sitting inside YouTube videos, Skala Blog turns it into a structured article: paste the video URL, get the transcription, and generate a publishable draft you can review and edit. The same goes for deep dives like this one; the written version is what people can quote, search, and check. More material lives at crazystack.com.br.

Source video