[{"data":1,"prerenderedAt":29},["ShallowReactive",2],{"nr-en-hassabis-fordert-globale-ki-aufsichtsbehoerde":3},{"slug":4,"title":5,"dek":6,"date":7,"time":8,"publishedAt":9,"updated":10,"updatedAt":10,"dateFmt":11,"updatedFmt":10,"kind":12,"tier":13,"author":14,"authorName":15,"topics":16,"tracker":22,"trackerLabel":23,"headlineStat":10,"image":24,"ogImage":25,"imageAlt":5,"csv":10,"minutes":26,"words":27,"html":28},"hassabis-fordert-globale-ki-aufsichtsbehoerde","Deepmind CEO Hassabis Calls for Global AI Watchdog Modeled on US Financial Regulator","Google Deepmind CEO Demis Hassabis has proposed a concrete regulatory framework: a new US agency modeled after FINRA to evaluate frontier AI models and potentially coordinate development slowdowns.","2026-07-14","16:34","2026-07-14T16:34:00+02:00","","July 14, 2026","news","standard","ideal-syka","Ideal Syka",[17,18,19,20,21],"AI regulation","governance","Deepmind","AGI","frontier models","\u002Feu-ai-act-fahrplan","EU AI Act","\u002Fnewsroom\u002Fimg\u002Fhassabis-fordert-globale-ki-aufsichtsbehoerde.webp","\u002Fog-nr\u002Fhassabis-fordert-globale-ki-aufsichtsbehoerde.en.png",3,521,"\u003Cp>Demis Hassabis, CEO of Google Deepmind, has published a detailed governance framework for advanced AI systems. His proposal: a new US standards body structured like the financial regulator FINRA that would develop evaluation protocols for frontier models.\u003C\u002Fp>\n\u003Ch2>The essentials\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Cstrong>Hassabis proposes FINRA-like agency\u003C\u002Fstrong> to develop evaluation protocols for frontier models – initially voluntary, later mandatory\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Agency could coordinate development slowdowns\u003C\u002Fstrong>, similar to what Anthropic recently considered\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Startups and academic research exempt\u003C\u002Fstrong> to avoid "regulatory capture"\u003C\u002Fli>\n\u003Cli>\u003Cstrong>International coordination planned\u003C\u002Fstrong>: The global community should follow suit and find consensus on critical points\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2>Timing is no coincidence\u003C\u002Fh2>\n\u003Cp>Hassabis' proposal arrives at a moment when the AI debate is accelerating. Shortly before, prominent AI researchers and economists issued an open letter warning of massive job losses from AI systems – Hassabis did not sign it, but argues similarly. His approach, however, is more concrete and less alarmist.\u003C\u002Fp>\n\u003Cp>The Deepmind CEO's reasoning is striking:\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>"Nobody in the world knows for sure what is going to happen from here, and even the experts disagree. When there is a large degree of uncertainty and the stakes are this high, proceeding with cautious optimism is the sensible and correct strategy."\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp>Hassabis reiterates an assessment he made in April: \u003Cstrong>according to Hassabis, AGI is likely within years\u003C\u002Fstrong>, with an impact ten times greater than the Industrial Revolution – and arriving ten times faster.\u003C\u002Fp>\n\u003Ch2>How the system would work\u003C\u002Fh2>\n\u003Cp>The proposed agency would be industry-funded and use regularly updated benchmarks. It could coordinate a development slowdown if necessary – a point showing Hassabis does not rule out a "pause."\u003C\u002Fp>\n\u003Cp>Crucially: \u003Cstrong>Non-frontier models from startups or academic institutions would be exempt\u003C\u002Fstrong>. This is a deliberate move to counter accusations that established firms use regulation to slow smaller competitors.\u003C\u002Fp>\n\u003Ch2>Experts disagree sharply\u003C\u002Fh2>\n\u003Cp>The debate over AI risks and AGI timelines is highly polarized within research. In December, Hassabis himself clashed publicly with Yann LeCun, who called the concept of general intelligence based on language models "complete BS." Hassabis disagreed publicly.\u003C\u002Fp>\n\u003Cp>Other researchers take a middle ground: Oriol Vinyals, co-lead of Gemini, sees today's models as strong in some areas but lacking true innovation capability. Deep learning pioneer Richard Sutton just announced his startup Oak Labs to tackle precisely this problem.\u003C\u002Fp>\n\u003Cp>Meanwhile, Deepmind co-founder Shane Legg considers "minimal AGI" possible as early as 2028.\u003C\u002Fp>\n\u003Ch2>What this means for companies globally\u003C\u002Fh2>\n\u003Cp>Hassabis' proposal is a strategic signal: Google is positioning itself as a proponent of structured AI governance – before regulatory pressure arrives externally. For companies worldwide, this could have several implications. First: if the US actually builds such an agency, international firms will face pressure to accept or adopt similar standards. Second: the focus on "frontier models" could mean smaller, specialized AI systems remain unregulated longer – an opportunity for mid-market players. Third: it remains unclear how the EU will respond and whether it will chart its own course – the EU AI Act may point in a different direction.\u003C\u002Fp>\n\u003Ch2>Sources\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fthe-decoder.com\u002Fdeepmind-ceo-hassabis-says-nobody-in-the-world-knows-what-happens-next-so-cautious-optimism-means-building-guardrails-now\u002F\">The Decoder\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.theverge.com\u002Ftech\u002F965270\u002Fgoogle-deepmind-demis-hassabis-global-ai-watchdog\">The Verge\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cem>Editorially owned by \u003Ca href=\"\u002Fen\u002Fautor\u002Fideal-syka\">Ideal Syka\u003C\u002Fa>. Sources and method: \u003Ca href=\"\u002Fen\u002Fredaktion\">Newsroom & method\u003C\u002Fa>. Tips and corrections: \u003Ca href=\"mailto:ai@i6eal.de\">ai@i6eal.de\u003C\u002Fa>.\u003C\u002Fem>\u003C\u002Fp>\n",1784042814265]