SEÁN Ó HÉIGEARTAIGH

Behavioral Archetype

THE CONVENER — Subject is the administrator who built the academic infrastructure of AI existential-risk research and then sat down at every table that infrastructure could reach. He was the founding Executive Director of Cambridge’s Centre for the Study of Existential Risk (CSER), working from 2013, a centre founded in 2012 by Martin Rees, Huw Price and Skype co-founder Jaan Tallinn. He helped write the proposal that won the University of Cambridge a £10 million Leverhulme grant for the Leverhulme Centre for the Future of Intelligence in 2015. He now directs the AI: Futures and Responsibility programme that joins the two centres, sits on the board of the Centre for the Governance of AI, was named in 2026 to the European Union’s AI Act Scientific Panel, advises the Survival and Flourishing Corp, whose primary client is Jaan Tallinn, and co-leads an AI safety and governance programme of the World Internet Conference’s committee on AI, an international organization headquartered in Beijing. The structural fact is the span: Cambridge, Brussels, Beijing and the donor’s advisory desk are one calendar. The method he has written down is relational and open about it: his own account of why he files evidence with Parliament is that it is “a good way to get established as a reputable, trustworthy expertise source and partner.” This profile scores the seat and the record, not the man: the hazard is the bridge, a single person who is simultaneously a fiduciary of the governance centre, a recommender in the donor’s grant round, an adviser to the donor’s corporation, an EU panel member and a co-lead in a Chinese-hosted governance programme, each of which is printed on a different page.

Essence Indicators

• Director of the AI: Futures and Responsibility Programme (AI:FAR) and Research Professor, University of Cambridge. The Leverhulme Centre for the Future of Intelligence (CFI) says “Professor Seán Ó hÉigeartaigh is Associate Director (Research Strategy) and the Programme Director for the AI:FAR research programme at the Leverhulme Centre for the Future of Intelligence (CFI)” (CFI, staff page); GovAI’s 2025 annual report calls him “the Director of the AI: Futures and Responsibility Programme, part of the Centre for the Future of Intelligence at the University of Cambridge, and a Research Professor at Cambridge” (GovAI, Annual Report 2025); CSER describes the programme as “a joint programme between CSER and the Centre for the Future of Intelligence (CFI) on the global risks and governance of advanced AI” (CSER, staff page)

• Founding Executive Director of CSER. CSER’s staff page: “Previously Sean was the founding Executive Director of CSER, leading its establishment with its founders from 2013 to 2015, and leading its major research projects from 2015 to 2020. He was also Acting Director of CSER from 2021-2023.” CSER’s account of its founding: “In 2012, together with Jaan Tallinn, the co-founder of Skype, and Huw Price, the Bertrand Russell Professor of Philosophy,” Martin Rees set out “to steer a small fraction of Cambridge’s great intellectual resources” toward the long-term future (CSER, staff page; CSER, Our story). The Future of Life Institute records that CSER’s next Director, Professor Matthew Connelly, began in July 2023 on a five-year post that its grant of $1,000,000 recommended in 2023 helped fund (FLI, grant to the University of Cambridge)

• Co-developer of the Leverhulme Centre for the Future of Intelligence. The University of Cambridge’s release of 3 December 2015: “an unprecedented £10 million grant from the Leverhulme Trust,” awarded “for a proposal developed with the Executive Director of the University’s Centre for the Study of Existential Risk (CSER), Dr Seán Ó hÉigeartaigh” (University of Cambridge, 3 December 2015). His earlier posts, per CRASSH: research programmes at the Future of Humanity Institute at Oxford from 2011 to 2015 and, “Prior to Cambridge,” the FHI-Amlin Collaboration on Systemic Risk, “a major academic-reinsurance industry partnership on catastrophic risk modelling” (CRASSH, people page)
• Board member of the Centre for the Governance of AI. GovAI’s 2025 annual report, Board section: “Seán is the Director of the AI: Futures and Responsibility Programme, part of the Centre for the Future of Intelligence at the University of Cambridge, and a Research Professor at Cambridge,” beside Tasha McCauley, Jeffrey Ding, Toby Ord and Helen Toner (GovAI, Annual Report 2025; GovAI, People)

• Member of the European Commission’s AI Act Scientific Panel. CFI, 2 July 2026: the Panel “appointed 60 independent experts,” and “Professor Seán Ó hÉigeartaigh… have joined the Panel for a 24-month term, renewable, in personal capacity”; the Panel “advises the AI Office and national authorities” on the AI Act (see European AI Office and EU AI Act) and on “assessment of the impacts and risks of General-Purpose AI (GPAI) models” (CFI, 2 July 2026). The panel’s creation and its 60 experts are also reported in the Commission’s announcement as carried by the ERA Portal Austria, dated 2 June 2026 (ERA Portal Austria, news item)

• Advisor to the Survival and Flourishing Corp (SFC) and a Recommender in the Survival and Flourishing Fund’s 2025 round. SFC’s team page lists him as “Advisor” and says “Our primary client is philanthropist Jaan Tallinn” (Survival and Flourishing Corp); the SFF-2025 recommendations page lists “Seán Ó hÉigeartaigh (Main Track)” among the Recommenders in “a grant-recommendation process for Funder Jaan Tallinn,” with “Total funding recommended: $34.33MM” (Survival and Flourishing Fund, SFF-2025)

• Co-lead of the AI Safety and Governance Program of the World Internet Conference’s Specialized Committee on AI. China Daily’s WIC pages describe him as “co-lead of the AI Safety and Governance Program of the WIC SC on AI” at a keynote in Beijing on 20 June 2025 and a seminar on 12 September 2025 chaired by Zeng Yi; the Beijing municipal government describes the World Internet Conference as “an international organization headquartered in Beijing” (China Daily, 21 June 2025; China Daily, 13 September 2025; Beijing Municipal Government, 13 July 2022)

On the Record

His words, in date order. Every quotation was matched to the page, paper or thread read. Where an item is under his byline for an institution, the post says so.

FLI post on the Leverhulme grant, 3 December 2015. Under his byline, the news release on the new centre: “The Centre is intended to build on CSER’s pioneering work on the risks posed by high-level AI and place those concerns in a broader context, looking at themes such as different kinds of intelligence, responsible development of technology and issues surrounding autonomous weapons and drones.” In the post he adds that CFI builds on “the generous support of Elon Musk, who helped massively boost this field with his (separate) $10M grants programme in January of this year,” and that “the Centre is funded for a full 10 years, and we will aim to build longer-lasting funding on top of this” (Future of Life Institute, 3 December 2015).

Paper with Stephen Cave, “An AI Race for Strategic Advantage: Rhetoric and Risks,” 2018. Abstract: “This paper assesses the potential risks of the AI race narrative and of an actual competitive race to develop AI, such as incentivising corner-cutting on safety and governance, or increasing the risk of conflict” (Cave and Ó hÉigeartaigh, AAAI/ACM Conference on AI, Ethics, and Society, 2018).

Advice and submissions, March 2019 (EA Forum). On a thread about the CSER and GovAI advice to the UN Secretary-General’s High-level Panel on Digital Cooperation, writing as “Sean_o_h” with the line “Disclosure: I am executive director of CSER,” he describes his method for open calls for evidence: “Submissions I’ve been involved in tend towards (a) summarising already published work (b) making sensible, noncontroversial recommendations (c) occasionally gently keeping the overton window open.” And the reason: “producing reputable input for them is a good way to get established as a reputable, trustworthy expertise source and partner.” He reports that the “CSER/CFI submissions to the House of Lords AI call for evidence was favourably received” and that “We received informal feedback that the submissions were influential” (EA Forum, comments on “CSER and FHI advice to UN High-level Panel on Digital Cooperation,” March 2019).

Paper with Whittlestone, Liu, Zeng and Liu, “Overcoming Barriers to Cross-cultural Cooperation in AI Ethics and Governance,” 2020. Abstract: “We argue that misunderstandings between cultures and regions play a more important role in undermining cross-cultural trust, relative to fundamental disagreements, than is often supposed,” and “academia has a key role to play in promoting cross-cultural cooperation on AI ethics and governance.” The author affiliations include the “China-UK Research Centre for AI Ethics And Governance, Beijing, China” and the Beijing Academy of Artificial Intelligence (Ó hÉigeartaigh et al., Philosophy & Technology, 2020).

Bulletin of the Atomic Scientists, 10 December 2025. “Stopping the Clock on catastrophic AI risk,” under his byline. The article body is behind the Bulletin’s premium wall; the reference list the page prints cites Anthropic’s May 2025 post “Activating AI Safety Level 3 Protections,” Sam Altman’s “Reflections,” “AI 2027,” the paper “Gradual Disempowerment” and the launch announcement of the political committee “Leading the Future” (The Bulletin, 10 December 2025).

Research listed on his staff pages, 2025 and 2026. CSER lists his working paper “The Most Dangerous Fiction: The Rhetoric and Reality of the AI Race” (SSRN, 2025), and CFI lists, with Yi-Ling Liu, Coleman Snell and Shakeel Hashim, “When the chips are down: Technology actors as power brokers in the US-China ‘AI Race’” (2026) (CSER, staff page; CFI, staff page).

Social Persona / Impression Management

Immediate impression: The administrator-scholar who writes the kind of paper a committee can adopt. His cited work is co-authored, cross-institutional and programmatic: a conference paper with a Cambridge colleague, a paper whose author list spans Cambridge, Beijing and Peking University, a 2026 paper with three co-authors. His own account of his method is that submissions are “sensible, noncontroversial,” with the occasional gentle push (EA Forum, March 2019).

Energy: Steady and organisational. The record is a chain of institutions: Oxford’s Future of Humanity Institute, CSER, the Strategic AI Research Centre, CFI, AI:FAR, GovAI’s board, the EU panel, the Beijing committee, each with a role title, none with a single-author book (CRASSH; CFI, staff page).

Impression management strategy: Reputation as an asset, stated plainly. His 2019 comment says that “It’s easier to justify investing time and resources in an org that’s been favourably featured in these processes as opposed to ‘random centre somewhere working on slightly unusual topics,’” and that the benefit of a submission is that it “allows people in government, including those already concerned with these issues, greater scope to engage with orgs like ours in more in-depth conversation” (EA Forum, March 2019). The method is visible in the roster: advisory work is listed first on the bio, in the order “OECD, Global Partnership on AI, United Nations, and UK Government up to Ministerial level” (CSER, staff page). Two limits are on the record: the posts for the Beijing programme give his role but no words of his, and the Bulletin article is paywalled, so his own words on the 2025 and 2026 work are not in this file.

• Social accounts and feeds. X: @S_OhEigeartaigh, LinkedIn and ORCID (0000-0002-2846-1576) are linked from his CFI staff page (CFI, staff page). EA Forum: writes as “Sean_o_h,” as above. Institutional accounts: CFI on Bluesky (cfi-cambridge.bsky.social) and CSER on X (@CSERCambridge) and Bluesky (cser.bsky.social), per their own pages. No personal blog, newsletter or Bluesky account of his own was found on his CSER or CFI pages; his Bulletin author page lists his Bulletin article (The Bulletin, author page).

Forensic Archetype Comparison

Psychometric Assessment

Observations of the documented public role from a record that is first-hand where it exists: his own forum comments, two papers, staff bios and a foundation post. Trait scores are style inferences, not claims about private character, and each says which exhibit it rests on.

Big Five (OCEAN):

Dark Triad:

MBTI: ENFJ (“The Protagonist”) — a convener who organises people across institutions and works through relationships and consensus papers. First-hand rationale: the CFI proposal that spanned four universities, the cross-cultural paper and his own account of why submissions build “partner” status (University of Cambridge, 3 December 2015; EA Forum, March 2019). This is a style inference from the documents, not a test result.

Threat Assessment

Score basis. ATK 6 is leverage: he sits on the Commission’s AI Act Scientific Panel, which “advises the AI Office and national authorities,” is a Recommender in a round that recommended $34.33 million, advises the donor’s corporation and directs a programme whose research is delivered to governments; it is not a 7 because the panel advises, the round’s recommendations are the Funder’s to act on, and he holds no enforcement, procurement or pre-release gate (CFI, 2 July 2026; SFF-2025). DEF 6 is insulation: a Cambridge research chair, a Leverhulme grant that “is funded for a full 10 years,” and philanthropic funding of CSER, none paid by a lab; held at 6 because the donor family overlaps across the centres, the grant round and the board he serves (University of Cambridge, 3 December 2015; FLI). HP 7 is durability: CSER (founded 2012, with a $1,000,000 FLI grant funding a five-year director post in 2023), CFI (a ten-year grant) and AI:FAR all outlive any holder of the seat (CSER, staff page). Against peers: Markus Anderljung (6/6/7, hazard 61) drafts instruments inside GovAI; Ó hÉigeartaigh sits on the EU panel, the SFF round and the Beijing committee as well as the same board, so he rates two points higher; Tasha McCauley (5/6/6, hazard 56) chairs GovAI’s US board but holds none of those seats; Helen Toner (6/8/8, hazard 80) leads a funded centre with executive authority. The 63 is defensible as a peer one step above Anderljung and well below Toner.

Alignment Analysis

Stated alignment: Reduce catastrophic risk from advanced AI by governance built across cultures and institutions, with the race framing treated as a risk. In his words: “the potential risks of the AI race narrative and of an actual competitive race to develop AI, such as incentivising corner-cutting on safety and governance”; “misunderstandings between cultures and regions play a more important role in undermining cross-cultural trust, relative to fundamental disagreements, than is often supposed” (Cave and Ó hÉigeartaigh, 2018; Ó hÉigeartaigh et al., 2020).

Observed alignment: Consistent with stated in the seats: he co-leads a Beijing-hosted governance programme with Chinese counterparts and a British co-chair, serves on an EU panel advising on general-purpose model risk, and keeps the Cambridge programme that joins existential-risk and AI-futures work. The overlap that needs the ledger is financial and fiduciary: he is a member of the board of GovAI, a Recommender in a round that gave GovAI $756,000 in total ($0 from his Main track), an advisor to a corporation whose primary client is the round’s Funder, and the director of a centre and programme created with money from that same donor family (GovAI, Annual Report 2025; SFF-2025; Survival and Flourishing Corp; CSER, Our story).

Gap assessment: The gap is structural, not personal. His published position is that the AI race is rhetoric with its own risks and that cooperation across regions is under-built; the committees he sits on are the ones that would design the cooperation, and the money that pays for the field is the money the round he sat on gave out. The record shows each tie and does not show that any of them shaped a recommendation: the SFF page says nothing on recusal and shows $0 from the Main track for GovAI, and the Beijing posts report his role but no remarks. What the record cannot settle is motive, and it never needs to. The finding is the position: one convener at the join of the field’s donor, its governance centre, a European panel and a Beijing committee, who has written down in public that a favoured submission “allows people in government… greater scope to engage with orgs like ours.”

Interest: Documented, never motive, never a verdict.

1. Position: the AI race is a narrative with its own risks, and cooperation, including with China, should be built (Cave and Ó hÉigeartaigh, 2018; Ó hÉigeartaigh et al., 2020). What adoption pays: a cooperation track needs conveners with counterpart access, and the roles the record shows are the ones he holds: co-lead of the WIC committee’s safety programme, co-director of the China-UK Research Centre for AI Ethics and Governance (the 2020 paper’s affiliation line), a Cambridge programme named in the Beijing seminar. No fee is documented; the programme is funded by the university and philanthropic sources. His best case: “We believe that academia has a key role to play in promoting cross-cultural cooperation on AI ethics and governance, by building greater mutual understanding.” Others who gain: the Chinese co-leads and institutions named in the WIC committee’s own pages (the Chinese Academy of Sciences’ Zeng Yi chairs the draft; Tencent, Weibo, MiniMax and other companies attend), and the Western labs and states that would otherwise face an adversarial race narrative.
2. Position: international and regional bodies should assess and govern frontier AI risk (the CSER and GovAI advice to the UN High-level Panel on Digital Cooperation, “A Proposal for International AI Governance,” March 2019, EA Forum; his EU panel seat, CFI). What adoption pays: every standing panel, commission or advisory body creates seats for the people who staff it; the EU panel is one he now holds, “in personal capacity,” for a renewable 24-month term, with no pay stated. The 2016 House of Commons submission he cites recommended “a standing Commission on Artificial Intelligence,” a body that, in his account, was followed by a government centre with “a senior CSER/CFI member on the board” (EA Forum, March 2019). His best case: open calls for evidence “are noisy processes,” and reputable input is how a small centre becomes a partner to people in government who are “already concerned with these issues.” Others who gain: regulators who need outside technical expertise on general-purpose models, and the frontier developers whose compliance the panels assess.
3. Position: philanthropic money should fund the AI-risk field at scale (his foundation post: “the Centre is funded for a full 10 years, and we will aim to build longer-lasting funding on top of this,” Future of Life Institute). What adoption pays: the centres he built and now directs; the donor whose round he sat on in 2025 recommended $34.33 million, and GovAI, on whose board he sits, is a $756,000 line in it, $696,000 of it a matching pledge (SFF-2025). His best case: his 2015 words, that the field needs “a long-term one in which the best talents can be expected to have lasting careers.” Others who gain: every grantee in the round, and the donor’s own priorities, named on the SFC page as its “primary client.”

Alliance Map

The Movement, Agency and Diaspora rows run at the same time, and that is the finding: the director who built Cambridge’s existential-risk centres sits in the same year on the board of the governance centre the donor family funds, in the donor’s grant round, on the donor corporation’s advisory desk, on an EU panel that advises the AI Office, and in a Beijing committee co-led with a Chinese Academy of Sciences scientist. Each tie is printed on its own page; none of the pages mentions another.

Convergent Drive Classification

Self-preservation: The seats overlap so that the loss of one is cushioned by the others: Acting Director of CSER until 2023, then AI:FAR director, then the EU panel in 2026, with the GovAI board and the SFC advisory desk running throughout. First-hand: CSER’s dates and CFI’s July 2026 announcement (CSER, staff page; CFI, 2 July 2026). Goal preservation: The thesis is stable from 2015 to 2025: AI risk deserves a long-lived research field, and the AI race is a framing that carries its own risk. First-hand: “the Centre is funded for a full 10 years” and the 2018 abstract (Future of Life Institute, 2015; Cave and Ó hÉigeartaigh, 2018). Resource acquisition: The resource is institutional money and seats: a £10 million Leverhulme grant won with his proposal, a donor family’s recommended grants, an EU panel seat and a Beijing committee role. First-hand: the University of Cambridge’s release and the SFF page (University of Cambridge, 3 December 2015; SFF-2025). Self-improvement: The role scale rises with each decade: project manager of an Oxford programme from 2011, founding Executive Director from 2013, programme director, Research Professor, panel member. First-hand: CFI’s and CRASSH’s bios (CFI, staff page; CRASSH).

Subject is not an AI system. The drives appear anyway, in the convener: the one who builds the room, funds the table and takes the seat, and who has said in public that the seat is how the next conversation starts.

Public footprint: CFI staff page · CSER staff page · X, @S_OhEigeartaigh · The Bulletin.

Related: GovAI; Jeffrey Ding; Tasha McCauley; Helen Toner; Markus Anderljung; Survival and Flourishing Fund; Jaan Tallinn; Future of Life Institute; Open Philanthropy; Longview Philanthropy; Revolving door graph.

Sources: CFI, Seán Ó hÉigeartaigh; CSER, Seán Ó hÉigeartaigh; CSER, Our story; CRASSH, Seán Ó hÉigeartaigh; GovAI, Annual Report 2025; GovAI, People; CFI, AI Act Scientific Panel, 2 July 2026; ERA Portal Austria, Commission expert bodies, 2 June 2026; Survival and Flourishing Corp; Survival and Flourishing Fund, SFF-2025; University of Cambridge, 3 December 2015; Future of Life Institute, grant to the University of Cambridge; Future of Life Institute, 3 December 2015; Cave and Ó hÉigeartaigh, 2018; Ó hÉigeartaigh et al., 2020; EA Forum, CSER and FHI advice to UN High-level Panel on Digital Cooperation; China Daily, 15 April 2025; China Daily, 21 June 2025; China Daily, 13 September 2025; Beijing Municipal Government, WIC inaugural ceremony; Singapore Conference on AI 2025; The Bulletin, Stopping the Clock on catastrophic AI risk; The Bulletin, author page