Andrea Miotti

Andrea Miotti is the Founder and CEO of ControlAI, the international AI-safety advocacy non-profit working through institutional channels — parliamentary briefings, written evidence, draft legislation, and weekly newsletter commentary — for binding regulation on the most powerful AI systems and an outright prohibition on the development of artificial superintelligence (see Person entry). He is tracked here as a Voice because his sustained public output — lead-authorship of ControlAI's flagship policy plan A Narrow Path, the weekly ControlAI Substack newsletter run co-bylined with Tolga Bilge, named UK Parliament written-evidence submission RAI0031, and recurring named commentary across TIME, The Guardian, Nature, BBC, and Sky News — has done as much as any single individual's to install the framing of frontier-AI governance and a superintelligence prohibition into UK and US parliamentary discourse and into mainstream press coverage of AI policy.

He is the corpus's first international AI-safety / Pause-and-Control adult Voice. Where the corpus's youth AI-safety Voice Sneha Revanur carries the case for grassroots-organising AI-policy advocacy through US state legislative campaigns and the open-letter circuit, the algorithmic-accountability Voices Joy Buolamwini and Matthias Spielkamp carry the case for near-term-harm regulation through audit science and EU AI Act trilogue work, and the worker-side Voice Daniel Motaung carries the case for content-moderator labour through the courts and the union, Miotti's voice carries the case that the largest end-state risks from artificial general intelligence and superintelligence are political-economy problems requiring binding international coordination, compute-threshold licensing, and a multi-decade moratorium on superintelligence development — the policy package A Narrow Path lays out and that ControlAI carries into UK, US, Canadian, and German parliamentary briefings. He fills the AI-safety / Pause-and-Control policy-track Voice slot structurally distinct from PauseAI's grassroots-organising pole and pairs with the corpus's ControlAI organisational anchor and A Narrow Path publication anchor as the named on-record voice carrying both forward into UK parliamentary and international press discourse.

Signature framings

Three framings in Miotti's public output have travelled beyond ControlAI's own materials and into the working language of the international AI-safety policy field.

• "A deeply political problem that will be solved with policy solutions, not just going back to the ivory tower." Miotti's signature founding-rationale framing is delivered in his Avoiding Extinction podcast conversation with Connor Leahy: "I founded ControlAI because I believe this is just, not just a technical problem … this is a deeply political problem that will be solved with policy solutions, not just going back to the ivory tower." The same conversation pairs the political-not-technical framing with the structural argument that "even with the best scientists working on this problem, we're not gonna make it out alive if we don't put rules in place, regulations in place that actually protect us." The formulation is the corpus's clearest single statement of the inside-game theory of change distinguishing ControlAI's Westminster parliamentary-briefing posture from the technical-AI-alignment research tradition out of which Miotti and ControlAI came.
• "AI systems smarter than all of you might combine are AI systems that will take over your country." Miotti's most-cited single superintelligence framing in the same conversation compresses the extinction-risk case into a national-security frame: "AI systems smarter than all of you might combine are AI systems that will take over your country." The line carries into the headline framing of A Narrow Path — that "artificial superintelligence threatens humanity with extinction" and "we do not know how to control AI vastly more powerful than us" — and is the rhetorical move ControlAI has used to make extinction-risk arguments legible inside parliamentary settings whose default vocabulary is defence-and-security rather than AI-alignment research. It is the framing that allowed cross-bench peers including the Rt Hon. the Lord Browne of Ladyton (former UK Defence Secretary, vice-chair of the Nuclear Threat Initiative) to sign on as ControlAI advisors.
• "There are only two times to react to an exponential — too early or too late." Miotti's exponential-timing framing, also delivered in the Avoiding Extinction podcast, is the urgency-of-action argument that anchors ControlAI's parliamentary-briefing case: there is no comfortable middle window from which to set rules on a technology improving at the rate the AI capability frontier is. The recurring weekly ControlAI Substack newsletter run that Miotti co-bylines with Tolga Bilge — including The Exponential on 6 March 2026 ("the difficulty of coding tasks AIs can reliably complete is still doubling every 4 months"; "Many experts believe superintelligence could arrive within just the next 5 years") — is the long-form vehicle through which Miotti carries the exponential-timing argument forward week by week into the empirical record on which the parliamentary briefings rest.

Public output and venues

Miotti's public-facing work runs across four overlapping channels.

• Lead-authorship of A Narrow Path. The single most load-bearing artefact of his voice is A Narrow Path: How to Secure Our Future, the long-form policy plan he lead-authored with Tolga Bilge, Dave Kasten, and James Newport and published on 19 March 2026 (anchored in corpus as pub-narrow-path). The plan's three-phase Safety / Stability / Flourishing framework — a twenty-year moratorium on superintelligence development, an international oversight architecture, and the eventual development of "safe-by-design" transformative AI under human control — built around compute-threshold licensing triggered at training runs above 10^25 FLOP, is the substantive policy package on which the rest of Miotti's public output rests, and the single artefact most frequently cited by other commentators when naming ControlAI's distinguishing contribution to the international AI-safety policy field.
• The weekly ControlAI Substack newsletter co-byline with Tolga Bilge. Miotti's most sustained public-output channel is the ControlAI Substack newsletter, where he is the co-bylined author with Tolga Bilge on a near-weekly cadence visible across the recent archive (The Exponential, 6 March 2026; New Poll: People Don't Want Superintelligence, 12 March 2026; We're Already Losing Control of AI Agents, 19 March 2026; Building the Coalition to Ban Superintelligence, 2 April 2026; How Can We Ban Superintelligence?, 23 April 2026; Why Do AI's Godfathers Warn of Extinction?, 30 April 2026; and so on through May 2026's Are We Close to an Intelligence Explosion?). The newsletter is the long-form vehicle through which the parliamentary-briefing programme's empirical case is updated week by week, and the channel through which Miotti's named-byline voice reaches the AI-safety policy audience between A Narrow Path-scale publications.
• Named UK Parliament written evidence and the cross-jurisdiction lawmaker-briefing programme. Miotti is the named co-author (with Steven Adler) of ControlAI's written-evidence submission RAI0031 to a UK parliamentary AI-regulation inquiry — the institutional-record artefact of his named-spokesperson role in the UK parliamentary track. He is also the public face of the parliamentary-briefing programme launched in September 2024, which by February 2026 had systematically briefed more than 250 lawmakers across the US, UK, Canada, and Germany and which produced the 11 December 2025 milestone of more than 100 cross-party UK parliamentarians publicly acknowledging extinction-level threat from advanced AI in a single binding-regulation campaign statement — covered in The Guardian on 8 December 2025 and earlier in TIME.
• Named commentary across the international press circuit. Miotti's expert commentary and op-eds have appeared, on ControlAI's own listing, across TIME, The Guardian, Nature, BBC, and Sky News. The cross-outlet press-record is the public-output channel through which the parliamentary-briefing programme's institutional work is translated into mass-press AI-policy coverage, and the venue through which Miotti is named as ControlAI's on-record public face by other commentators citing his and the organisation's work. Long-form podcast appearances (the Avoiding Extinction conversation with Connor Leahy on ControlAI's own podcast track) complement the press circuit and are the venues at which his signature framings have been delivered in their fullest form.

Why this is a Voice entry

A Voice entry is created here, rather than additional structure on the Person entry, because Miotti's public-facing output is itself the load-bearing object the corpus needs to track: the working international vocabulary of binding AI regulation, compute-threshold licensing, and a twenty-year superintelligence moratorium — the three-phase Safety / Stability / Flourishing framework he lead-authored, the political-not-technical framing that he founded ControlAI on, and the exponential-timing framing carried weekly into the ControlAI Substack newsletter — is the framing he has installed into UK and US parliamentary discourse and international press coverage of AI policy over the three-year arc of ControlAI's public work since 2023. The corpus's AI-safety / Pause-and-Control Voice slot — the policy-track inside-game pole structurally complementary to the grassroots-organising side anchored in PauseAI and to the youth-side anchored in voice-sneha-revanur — carried no adult Voice before this entry; this entry gives that pole its first first-person voice. Affiliation, prior employment, and biographical structure are recorded on the linked Person entry per the corpus's Person/Voice split.