The Eight-Year Influence Operation

From one London startup to a US BAN on Open AI

From a Labour-adjacent London startup called Brixton Endeavours to a bipartisan American ban on open-source AI. This is what an institutional pipeline looks like when it works.

---

What this piece is

Over the past two weeks I published four investigations documenting different phases of one operation:

1. Piece 1 — The CCDH Rebrand Nobody Noticed documented the September 8, 2026 Coxon launch and the Center for Countering Digital Hate’s twenty-nine-day rebrand into “Center for Countering Digital Harm.”

2. Piece 2 — Bernie’s Anti-Oligarch Bill Was Written by the Oligarchs documented the Open Philanthropy-Anthropic funder graph behind the Sanders-Casar Ban Superintelligence bill.

3. Piece 3 — Obama Just Endorsed a Manufactured Panic documented the September 10 Obama endorsement and the Obama-alumni policy shop running Anthropic’s Washington operation.

4. Piece 4 — The AI Bill Bernie Sanders Introduced Was Drafted in London documented that ControlAI, a two-person London lobby shop, drafted both the UK Sobel bill and the Sanders-Casar bill — and that its only publicly named funder led Anthropic’s Series A.

This piece is the connective tissue. It puts the whole eight-year architecture in one place, with primary-source verification, so a reader, a reporter, or a lawyer can see the operation whole.

The thesis is unglamorous: a small circle of London-based advocacy organizations — CCDH founded 2018, ControlAI incorporated 2023, both operating as lobbying and communications entities under UK company law — have produced the analytical framework and, in ControlAI’s case, the legislative text now driving federal regulation of speech and AI in the United States. The through-line personnel are Labour-adjacent. The through-line funder ecosystem is Effective-Altruism-adjacent. The through-line beneficiary is the incumbent frontier AI lab whose Series A the same funder led.

None of this requires anyone to have secretly coordinated with anyone. What it required was that a small institutional pipeline was allowed to operate for eight years without disclosure — and to reach the point where a US Senator introduces a bill drafted by a foreign lobby shop and the American press covers it as domestic policy.

The master timeline

The nine-station institutional pipeline, from Brixton Endeavours to Sanders-Casar + Sobel.

Every entry that follows is anchored to a primary source. Where a claim cannot be verified from a primary source, I say so.

2018 — Foundation

October 19, 2018. A UK company is incorporated as “BRIXTON ENDEAVOURS LIMITED,” Companies House number 11633127. Its founding director is Morgan McSweeney. McSweeney was the recently resigned Chief of Staff of former Prime Minister Keir Starmer. That is a matter of Companies House record — his director appointment shows “Appointed on 19 October 2018 — Resigned on 6 April 2020,” Irish nationality, resident England.

Two months earlier, in July 2018, the House of Commons DCMS Committee chaired by Conservative MP Damian Collins publishes its Interim Report on “Disinformation and ‘fake news’”. It calls for statutory regulation of platforms.

2019 — Naming and framing

February 18, 2019. Collins’s DCMS Committee publishes its Final Report on Disinformation and ‘fake news’, containing a section titled “Online harms and regulation.” The report explicitly awaits and endorses the Government’s forthcoming Online Harms White Paper.

April 8, 2019. The UK Government publishes the Online Harms White Paper. It contains one operative sentence: “The White Paper proposes establishing in law a new duty of care towards users, which will be overseen by an independent regulator.” That is the structural genome of everything that follows.

August 30, 2019. Brixton Endeavours Limited is renamed “CENTER FOR COUNTERING DIGITAL HATE LTD” at Companies House (a fact recorded in the company’s previous-names table). The organization launches publicly under founder Imran Ahmed, who — per multiple secondary sources — had worked for Labour MPs Andy Slaughter and Hilary Benn, ran communications for Angela Eagle‘s July 2016 Labour leadership campaign, and was working as a Labour advisor in Portcullis House in March 2017.

2020 — Building the board

May 4, 2020. Simon James Clark appointed director of CCDH (Companies House officers list).

July 20, 2020. Damian Collins MP — the sitting chair of the DCMS Committee that pushed the “online harms” framework into government — is appointed to the CCDH board as an unpaid director. His statutory director appointment is registered at Companies House; his declaration to the House Register of Members’ Financial Interests reads: “From 20 July 2020 until further notice, unpaid board member of the Center for Countering Digital Hate, a not for profit NGO headquartered in London… 2 hrs per month.”

2021 — The pipeline delivers

March 24, 2021. CCDH publishes the “Disinformation Dozen” report, naming twelve Americans — me among them — as responsible for “up to 65%” of anti-vaccine content on Meta platforms.

February 4, 2021 (seven weeks earlier). The Institute for Strategic Dialogue publishes “The yoga world is riddled with anti-vaxxers and QAnon believers,” authored by ISD UK’s Cécile Simmons, which names me and greenmedinfo.com. The UK infrastructure for identifying the American targets is in place before the American report is published.

July 15, 2021. In the White House briefing room, Surgeon General Vivek Murthy and Press Secretary Jen Psaki deploy CCDH’s central finding verbatim: “there’s about 12 people who are producing 65 percent of anti-vaccine misinformation on social media platforms.” — record copy filed in the Supreme Court’s cited-URLs collection for Murthy v. Missouri, No. 23-411.

July 16, 2021. Follow-up White House briefing — Psaki confirms “we’re talking about these individuals.”

September 9, 2021. Damian Collins is appointed chair of the Joint Committee on the Draft Online Safety Bill. At the session where CCDH’s CEO Imran Ahmed appears as a witness, Collins declares “a relevant non-pecuniary interest as a member of the Center for Countering Digital Hate’s advisory board.” CCDH’s written evidence is cited at multiple points in the Committee’s final report.

September 9, 2021. Ahmed himself is appointed a director of CCDH at Companies House. His entry lists “Place of residence: United States.”

December 14, 2021. The Joint Committee’s Final Report is published. Collins quoted: “we need to call time on the Wild West online.”

2022 — Bridge years

February 16, 2022 – April 19, 2023. Additional CCDH directors appointed: Matthew Adam Gould, David Alan Rich, Ian Alexander Russell — completing the eight-person statutory board that counterhate.com does not publish on its own website.

July 22, 2022. Damian Collins resigns his CCDH directorship upon appointment as UK Under-Secretary of State for Tech and Digital Economy.

August 22, 2023. A new UK company is incorporated: “SECURE FUTURE RESEARCH LTD”, Companies House number 15088415, registered at 201-211 Borough High Street, London SE1 1JA. Its filed nature of business is SIC code 70210 — “Public relations and communications activities.” This is the corporate shell of what will publicly become ControlAI.

October 28, 2022. Damian Collins is re-appointed a CCDH director. That appointment is still active.

October 26, 2023. The UK Online Safety Act 2023 receives royal assent, codifying: a residual “material risk of significant harm” category (s.60(2)(c)); penalties up to “£18 million” or “10% of the person’s qualifying worldwide revenue” (Sch. 13 para. 4); criminal liability for named senior managers (s.110); and — critically — extraterritorial reach explicitly applying that criminal liability “to acts done by an individual in the United Kingdom or elsewhere” (s.205(2)).

November 7, 2023. Nicolas Michel Pierre Moës resigns as a director and Person with Significant Control of ControlAI, seventy-eight days after incorporation (filing history). The two remaining PSCs are Andrea Miotti (Italian, b. Nov 1996) and Eva Behrens (German, b. Feb 1995), each with 25-50% voting rights and the right to appoint or remove directors.

2024–2025 — Personnel migration

January 2024. Portions of a leaked CCDH “Portfolio Planning Monthly Meeting” note surface (documented in later reporting). One line, attributed to Ahmed: “RFK — black ops being set up to look at RFK.”

April 2024 – October 2025. CCDH publishes at least sixteen AI-tagged research reports building the case for a federal AI regulator, including reports on Bard, ChatGPT, Midjourney, Grok, Facebook, DALL-E, and voice-cloning tools.

October 2025. ControlAI’s founder Jaan Tallinn personally signs a Future of Life Institute statement calling for “a prohibition on the development of superintelligence, not lifted before there is broad scientific consensus that it will be done safely and controllably.” The statement lays out the exact policy the Sanders-Casar and Sobel bills implement one year later.

October 13, 2025. Ofcom issues the first-ever Online Safety Act penalty — against a US company, imageboard site 4chan. The £20,000 fine escalates by £100 per day. It is subsequently increased to £520,000 (Columbia Global Freedom of Expression case note). 4chan and Kiwi Farms respond by suing Ofcom in US federal court, claiming its enforcement “violated Americans’ right to freedom of speech.”

November 2025. Anthropic hires Maxwell Young as head of policy communications; his prior role was aide to Senate Democratic Leader Chuck Schumer.

December 9, 2025. House Democratic Leader Hakeem Jeffries announces the Democratic Commission on AI and the Innovation Economy. Democrats only. Co-chairs: Ted Lieu, Josh Gottheimer, Valerie Foushee.

2026 (January–August) — Positioning

January 2026. A new X account, @JacobCoxon, is created. It posts ten tweets and gains thirteen followers over the next eight months.

February 12, 2026. Anthropic announces a $20 million corporate contribution to Public First Action, a 501(c)(4). Co-founder Jack Clark tells the New York Times the company had spent “at least five months actively working on a super PAC strategy.”

March 11, 2026. Axios reports that Sarah Heck has become Anthropic’s Head of Public Policy, succeeding co-founder Jack Clark and reporting directly to Anthropic President Daniela Amodei. Heck’s prior public-record role: Director of Global Engagement at the Obama White House National Security Council, ending in February 2017 when she left for Stripe. That means Anthropic’s chief policy officer during the operation is an Obama NSC alumnus.

March 2026. ControlAI names Connor Leahy — co-founder of Conjecture, the AI-safety startup where Andrea Miotti previously served as Head of Strategy and Governance — as US Executive Director. (Wikipedia records: “In March 2026 Connor Leahy… was appointed US Executive Director.”)

May 14, 2026. The Gates Foundation announces a $200 million partnership with Anthropic — the Foundation’s largest publicly disclosed direct partnership with any frontier AI lab.

July 22, 2026. Anthropic announces a second $20 million contribution to Public First Action. Total: $40 million.

July 23, 2026. The AI Kill Switch Act (H.R. 9917) is introduced by Rep. Ted Lieu and Rep. Jared Moran.

August 4, 2026. Anthropic announces former California Supreme Court Justice Mariano-Florentino “Tino” Cuéllar as its first-ever Chief Global Affairs Officer.

August 10, 2026. CCDH’s board passes a formal company-name-change resolution (RES15) (filing history).

August 12, 2026. Companies House issues the certificate. Center for Countering Digital Hate Ltd. is legally renamed Center for Countering Digital Harm Ltd. The single-letter change tracks precisely to the “harm” vocabulary of the OSA — the statutory ground on which future enforcement will proceed.

August 26, 2026. Bill Gates publishes “The turbulent AI era is here” — including the operative sentence: “If someone had a credible plan for slowing down AI advances globally, I would likely support it.”

2026 (September) — Ignition

September 2, 2026. Anthropic researcher Joe Benton announces his move from Anthropic to METR. His personal website: “I also spent time at the UK Frontier AI Taskforce (now the UK AI Security Institute), helping to set it up in its early days.” The bridge between the US AI-safety funder graph and the UK’s own AI regulator is publicly named.

September 3, 2026. Senator Bernie Sanders and Representative Greg Casar announce the Ban Artificial Superintelligence Act. The summary PDF specifies: “a new cabinet-level federal agency,” “the corporate death penalty,” individual penalties of “not more than 20 years in prison,” and “international agreements” to ban superintelligence globally. The release does not name any outside drafter.

September 8, 2026.

• 3:13 PM Eastern. The Wall Street Journal publishes “What to Know About Anthropic’s Planned IPO.”

• 7:46 PM Eastern. WSJ publishes “Exclusive: Anthropic Researcher Jacob Coxon Quits Over ‘Out-of-Control’ AI Fears.”

• 8:04 PM Eastern. Coxon posts his resignation tweet. Eighteen minutes after WSJ.

Also on September 8, 2026: In London, MP Alex Sobel introduces the Artificial Superintelligence Bill at first reading. From Hansard: “I also would like to thank ControlAI for its support with the Bill.”

Also on September 8, 2026: TIME publishes Andrea Miotti of ControlAI acknowledging that the group “wrote the U.K. bill and consulted on the U.S. one.”

September 9, 2026. Wired publishes: “In an unusual move, the parliamentary bill was drafted by the nonprofit ControlAI, which is backed by Skype founding engineer and former DeepMind investor director Jaan Tallinn.” Also: “Conjecture founder Connor Leahy now leads ControlAI’s US arm, which consulted on Bernie Sanders and Greg Casar’s bill… in addition to having briefed nearly 200 congressional offices and more than 20 members of the Senate and House of Representatives.”

September 10, 2026, 6:00 PM Eastern. At a closed-press DCCC fundraiser at the New-York Historical Society, House Minority Leader Hakeem Jeffries asks Barack Obama about AI. Obama responds — per the September 13 New York Times account: “AI is moving very fast in private hands. Once you are speaker, I would strongly urge that the Democrats put together a framework.” Jeffries explicitly references the Coxon resignation.

September 11, 2026. The New York Times opinion section publishes Stephen Witt’s “This Is Really Bad” — a 21-minute prominent-placement audio op-ed endorsing both the Sanders-Casar bill and the Lieu-Moran Kill Switch Act.

September 11, 2026. Seventy-plus UK MPs and peers send a letter to Prime Minister Andy Burnham demanding adoption of ControlAI’s Sobel bill and an international ban.

September 12, 2026. Anthropic CEO Dario Amodei publishes “We Must Pace the Frontier,” demanding that governments “require every frontier AI firm to embed third-party evaluators ‘such as METR’ with employee-level access” — the essay names METR by name.

September 13, 2026. Speaker Mike Johnson (Republican) publicly states that AI leaders must meet with Congress on safety laws after “end of humanity” warnings. He resists the moratorium — but accepts the frame. Bipartisan capture of the frame is now complete, twelve days after Sanders’ announcement.

The mechanism map: OSA → Sanders-Casar

The two bills are not identical, but they use the same architectural template. Here is the correspondence, verified from primary statutory text on both sides:

Table Links: s.60(2)(c): Sch. 13 para. 4: .s110: s.205(2)

The Online Safety Act 2023 and the Ban Superintelligence Act use the same architecture.

The honest thesis is methodological lineage, not textual cloning: define the harm by open-ended outcome, hand the boundary-drawing to the executive or a purpose-built regulator, apply criminal exposure and revenue-based fines, project the standard outward beyond the jurisdiction, back it with an international agreement. That template is now visible in three places — the OSA (enacted), the Sobel bill (introduced, awaiting print), and the Sanders-Casar bill (announced, awaiting introduction).

Who moves between the layers

The London-based advocacy layer, the US frontier-AI-lab layer, and the American political layer are not three separate ecosystems. The following personnel data is on the public record:

• Morgan McSweeney — founding director of what became CCDH; now Chief of Staff to UK Prime Minister Keir Starmer. Companies House record: “Appointed on 19 October 2018 — Resigned on 6 April 2020.”

• Damian Collins MP — Conservative; chair of the DCMS Committee (2016–19) and the Joint Committee on the Draft Online Safety Bill (2021); CCDH director 2020-22 and again from 28 October 2022 to present; ministerial post at DCMS in between.

• Imran Ahmed — CCDH founder and CEO; per multiple secondary sources, staffer to Labour MPs Andy Slaughter and Hilary Benn, communications chief for Angela Eagle’s 2016 Labour leadership campaign. Current Companies House listing: “Place of residence: United States.”

• Chloe Colliver — now Head of Industry Developments for Online Safety at Ofcom, per her CERRE biography; previously “led the Institute for Strategic Dialogue’s digital analysis unit.” ISD published the article naming me and greenmedinfo.com in February 2021 — seven weeks before the Disinformation Dozen.

• Andrea Miotti — Head of Strategy and Governance at Conjecture (June 2022 to October 2023 founding of ControlAI); ControlAI founder and 25-50% PSC.

• Connor Leahy — Conjecture co-founder; ControlAI US Executive Director (March 2026).

• Gabriel Alfour — Conjecture co-founder; per ControlAI’s own about page, “helped Andrea Miotti found ControlAI.”

• Jaan Tallinn — ControlAI’s sole publicly named funder (per Guardian and Wired); board observer at Anthropic; led Anthropic’s Series A (per Wikipedia).

• Sarah Heck — Anthropic Head of Public Policy since March 2026; formerly Director of Global Engagement at the Obama NSC.

• Anna Makanju — OpenAI VP of Global Affairs; per her Obama Foundation biography, “eight years in the Obama/Biden administration.”

• Terah Lyons — Policy Advisor to the Obama White House CTO; per her own site, “co-directed the White House’s Future of Artificial Intelligence Initiative.”

• Joe Benton — ex-Anthropic Scalable Oversight team; joined METR September 2, 2026; per his own site, previously helped set up the UK Frontier AI Taskforce (now UK AI Security Institute).

• Maxwell Young — Anthropic head of policy communications since November 2025; prior role, aide to Senate Democratic Leader Chuck Schumer.

• Mariano-Florentino Cuéllar — Anthropic’s first Chief Global Affairs Officer since August 4, 2026; former California Supreme Court Justice.

The pattern: UK-institutional advocacy origin → US frontier-AI-lab public-affairs seat → Democratic-party operational adjacency. No single crossing proves coordination. The whole pattern is what proves an operating pipeline.

What the money says

• Anthropic → Public First Action: $40 million across two contributions ($20M February 12, 2026 + $20M July 22, 2026).

• Dario Amodei personal → Public First: $1 million on May 4, 2026.

• Anthropic 2025 lobbying: $3.1 million (per Anthropic’s own Senate LDA filings).

• Anthropic H1 2026 lobbying: $3.5 million-plus.

• Anthropic employee 2026-cycle contributions: $11.98 million (versus $220,180 in the entire 2023-24 cycle — a 54x increase).

• Gates Foundation → Anthropic: $200 million partnership announced May 14, 2026.

• Open Philanthropy → Conjecture (ControlAI leadership’s prior organization): $1.15 million across four grants, 2022-25 (per the Open Philanthropy grants archive).

• Jaan Tallinn total documented AI-safety giving: ~$37.98 million across 34 donees since 2012, per the Vipul Naik donations tracker.

What Ofcom is actually doing

Because the Online Safety Act has been enforced for one year, we now know what “duty of care” enforcement looks like in practice. As of October 2025 through February 2026:

• Ofcom’s first-ever OSA penalty went to a US company: 4chan Community Support LLC, ultimately £520,000 (case note).

• 4chan and Kiwi Farms sued Ofcom in US federal court arguing Ofcom’s enforcement “violated Americans’ right to freedom of speech.”

• Additional open OSA investigations target US-form entities including Score Internet Group LLC, plus Krakenfiles, Nippyspace, Im.ge, and Yolobit.

The lesson: the regulator with vaguely-defined “harm” powers, extraterritorial reach, and revenue-based fines has used them against US companies within the first year of enforcement. That is not a hypothetical prediction. That is the operating record of the exact template now proposed for US federal AI regulation.

What the CCDH-to-Harm rename actually signifies

CCDH renaming itself Center for Countering Digital Harm — twenty-seven days before the Coxon operation — is legible only against this template. “Harm” is the statutory vocabulary of the OSA. It is the vocabulary Ofcom uses. It is the vocabulary the Sobel bill uses (“serious damage to the security of the United Kingdom”). It is the vocabulary the Sanders-Casar bill’s federal AI agency will use.

An organization does not change its legal name for cosmetic reasons. A UK company limited by guarantee changes its legal name by board resolution (RES15) that is minuted, dated, and filed at Companies House. The name change is a governance decision. Its timing — twenty-seven days before the operation that would give Sanders his platform — is either coincidence or positioning. I do not claim to know which. What I can document is that the same UK NGO whose “Disinformation Dozen” framework the Biden White House deployed against six Americans in 2021 has renamed itself into the exact statutory vocabulary the AI-regulation bills use in 2026. That fact is on the public record and no reporter has yet asked CCDH to explain it.

What is not claimed

I do not claim that CCDH, ControlAI, Anthropic, Bernie Sanders, Greg Casar, Alex Sobel, Barack Obama, or the White House coordinated the events described. I have no evidence of that. What I claim is a documented institutional pipeline: same UK NGO ecosystem across three different phases (2018 online harms framework → 2021 Disinformation Dozen → 2026 AI-ban bills), same Effective-Altruism-adjacent funder ecosystem in the middle, same Anthropic beneficiary structure at the end, same Obama-alumni policy layer moving the frame through Washington.

I do not claim the OSA and the Sanders-Casar bill are textually identical. They are not. The claim is architectural: open-ended definition, delegation to a purpose-built regulator, revenue-based fines, executive criminal liability, extraterritorial reach, backed by an international agreement.

I do not claim Chloe Colliver personally named me at the Institute for Strategic Dialogue. She did not — the February 4, 2021 ISD article is bylined ISD UK’s Cécile Simmons. What I claim is that Colliver led ISD’s digital analysis unit before moving to Ofcom, where she is now Head of Industry Developments for Online Safety — an ISD → Ofcom pipeline that is on the public record.

Every date, quotation, and financial figure in this piece is sourced inline. Corrections to info@greenmedinfo.com and I will issue a same-day public correction if any fact is wrong.

What comes next

Two things are on the near-term calendar:

1. November 13, 2026 — second reading of the Sobel Artificial Superintelligence Bill in the UK House of Commons. The bill has not yet been printed. When it is printed, we will finally have a formal comparison of its clause structure with the Sanders-Casar summary.

1. The Anthropic S-1 — filed confidentially June 1, 2026, expected to become public in the coming weeks; the IPO reportedly targeted at up to $100 billion raised on a $2 trillion valuation.

The window in which the American public, US Members of Congress, UK parliamentarians, and constitutional lawyers can absorb what has been documented above is the window between now and those two events. This is what an institutional pipeline looks like when it works — and what a small coalition of primary-source journalists, plaintiffs’ lawyers, and civil-society organizations still has time to interrupt.

The organizing already underway: Restore the First — the coalition that formed to end the last iteration of this pattern and is now positioned to prevent the next. The full constitutional argument, including why licensing the tools of speech (AI models included) is prior restraint under any honest reading of the First Amendment, is here.

---

Sources cited in this piece

Companies House (UK):

• CCDH / Digital Harm Ltd overview: https://find-and-update.company-information.service.gov.uk/company/11633127

• CCDH officers list: https://find-and-update.company-information.service.gov.uk/company/11633127/officers

• CCDH filing history: https://find-and-update.company-information.service.gov.uk/company/11633127/filing-history

• ControlAI / Secure Future Research Ltd: https://find-and-update.company-information.service.gov.uk/company/15088415

• ControlAI PSC register: https://find-and-update.company-information.service.gov.uk/company/15088415/persons-with-significant-control

• ControlAI filing history: https://find-and-update.company-information.service.gov.uk/company/15088415/filing-history

UK Parliament & law:

• Online Safety Act 2023: https://www.legislation.gov.uk/ukpga/2023/50

• Online Harms White Paper: https://www.gov.uk/government/consultations/online-harms-white-paper

• DCMS Committee Final Report on Disinformation: https://publications.parliament.uk/pa/cm201719/cmselect/cmcumeds/1791/1791.pdf

• Joint Committee on the Draft Online Safety Bill Report: https://publications.parliament.uk/pa/jt5802/jtselect/jtonlinesafety/129/129.pdf

• Joint Committee Formal Minutes: https://committees.parliament.uk/publications/8257/documents/84343/default/

• Sobel bill page: https://bills.parliament.uk/bills/4288

• ACOBA advice letter (Damian Collins ↔ CCDH): https://www.gov.uk/government/publications/collins-damian-parliamentary-under-secretary-of-state-for-tech-and-digital-economy-at-the-department-for-digital-culture-media-and-sport-acoba/advice-letter-damian-collins-member-of-the-advisory-board-center-for-countering-digital-hate

US legal and government record:

• Murthy v. Missouri opinion (No. 23-411): https://www.supremecourt.gov/opinions/23pdf/23-411_3dq3.pdf

• Murthy Respondents’ Brief: https://www.supremecourt.gov/DocketPDF/23/23-411/299644/20240202144405984_2024-02-02%20-%20Murthy%20v.%20Missouri%20-%20Brief%20of%20Respondents%20-%20Final%20with%20Tables.pdf

• Psaki/Murthy July 15, 2021 briefing (record copy): https://www.supremecourt.gov/opinions/URLs_Cited/OT2023/23-411/23-411-13.pdf

• Psaki July 16, 2021 briefing: https://www.whitehouse.gov/briefing-room/press-briefings/2021/07/16/press-briefing-by-press-secretary-jen-psaki-july-16-2021/

• Sanders-Casar release: https://www.sanders.senate.gov/press-releases/news-sanders-casar-introduce-legislation-to-ban-artificial-superintelligence-and-temporarily-pause-advanced-ai-development/

• Sanders-Casar summary PDF: https://www.sanders.senate.gov/wp-content/uploads/Ban-Artificial-Superintelligence-Act-Release-Summary.pdf

• H.R. 9917 (Kill Switch): https://www.congress.gov/bill/119th-congress/house-bill/9917/text

• Jeffries AI Commission: https://jeffries.house.gov/2025/12/09/leader-jeffries-announces-new-house-democratic-commission-on-ai-and-the-innovation-economy/

Press:

• Wired UK ASI (Sept 9, 2026): https://www.wired.com/story/uk-lawmakers-are-scrambling-in-response-to-ai-summer-of-chaos/

• TIME UK/US ban (Sept 8, 2026): https://time.com/article/2026/09/08/ban-superintelligence-ai-uk-us-lawmakers/

• Guardian, Burnham letter: https://www.theguardian.com/technology/2026/sep/11/mps-urge-andy-burnham-block-artificial-superintelligence-asi

• WSJ Coxon exclusive: https://www.wsj.com/tech/ai/anthropic-researcher-jacob-coxon-quits-out-of-control-ai-fears

• NYT Obama piece (Sept 13, 2026): https://www.nytimes.com/2026/09/13/us/politics/obama-democrats-ai.html

• NYT Witt opinion (Sept 11, 2026): https://www.nytimes.com/2026/09/11/opinion/ai-safety-threat-technology.html

• Axios (Anthropic DC): https://www.axios.com/2026/03/11/anthropic-dc-presence

• Reuters (Anthropic Public First Action, Feb 12): https://www.reuters.com/legal/government/anthropic-donate-20-million-us-political-group-backing-ai-regulation-2026-02-12/

• Reuters (Cuéllar): https://www.reuters.com/world/anthropic-names-global-affairs-chief-tackle-ai-policy-trump-tensions-persist-2026-08-04/

• Reuters (first OSA fine): https://www.reuters.com/legal/litigation/britain-issues-first-online-safety-fine-us-website-4chan-2025-10-13/

• ISD, “The yoga world is riddled with anti-vaxxers and QAnon believers” (Feb 4, 2021): https://www.isdglobal.org/digital-dispatch/the-yoga-world-is-riddled-with-anti-vaxxers-and-qanon-believers/

Organizational and personnel:

• ControlAI About: https://controlai.org/about

• ControlAI Statement: https://controlai.org/statement

• ControlAI Designing the DIP: https://controlai.org/designing-the-dip

• ControlAI UK bill page: https://controlai.org/uk-asi-bill

• A Narrow Path: https://www.narrowpath.co/

• CERRE Chloe Colliver bio: https://cerre.eu/biographies/chloe-colliver/

• Joe Benton personal site: https://joejbenton.com/

• Wikipedia Jaan Tallinn: https://en.wikipedia.org/wiki/Jaan_Tallinn

• Wikipedia ControlAI: https://en.wikipedia.org/wiki/ControlAI

• Wikipedia Connor Leahy: https://en.wikipedia.org/wiki/Connor_Leahy

• Vipul Naik donations tracker: https://donations.vipulnaik.com/donor.php?donor=Jaan+Tallinn

• Ofcom enforcement: https://www.ofcom.org.uk/enforcement

• Columbia Global Freedom of Expression 4chan case: https://globalfreedomofexpression.columbia.edu/cases/the-case-of-the-ofcom-investigation-into-4chan/

• Obama Foundation Democracy Forum 2023 (Makanju): https://www.obama.org/democracy-forum-2023/anna-makanju/

• Terah Lyons site: https://www.terahlyons.com/

• Wikipedia Imran Ahmed: https://en.wikipedia.org/wiki/Imran_Ahmed_(strategist)

Prior published pieces in this series:

• Piece 1: https://sayerji.substack.com/p/anthropics-ai-whistleblower-and-the

• Piece 2: https://sayerji.substack.com/p/bernies-anti-oligarch-bill-was-written

• Piece 3: https://sayerji.substack.com/p/obama-just-endorsed-a-manufactured

• Piece 4: https://sayerji.substack.com/p/exclusive-the-ai-bill-bernie-sanders

Coalition:

• Restore the First: https://restorethe1st.com/

• Restore the First — Why AI is a 1A fight: https://restorethe1st.com/why-ai-is-1a.html

---

Sayer Ji is a plaintiff in Finn v. Global Engagement Center (dismissed with prejudice July 28, 2026 on a shotgun-pleading defect, without reaching standing or the merits) and the founder of GreenMedInfo. He was named on the Center for Countering Digital Hate’s March 2021 “Disinformation Dozen” report.