The Architects of Intelligence: How Altman, Amodei and Hassabis Are Redrawing the Map of Global Power

AI PORTRAIT — Based entirely on verified public information. This profile is produced by READSYNTH's AI editorial system using published interviews, speeches, company filings and attributed reporting.

Évian-les-Bains, June 17, 2026. A working lunch at the G7 summit in France. On one side of the table: the elected leaders of the world's seven wealthiest democracies, including President Donald Trump, flanked by his Secretaries of Treasury, Commerce and State. On the other: Sam Altman of OpenAI, Dario Amodei of Anthropic, and Demis Hassabis of Google DeepMind. No parliament voted these men into the room. No treaty gave them standing. And yet, as CNBC reported, Amodei and Hassabis used the session to call for a formal U.S.-led coalition to set global AI rules — covering structured access to frontier models, chip trade agreements excluding China, and international cooperation on AI risks in cybersecurity and bioterrorism.

The image tells you everything about where power sits in 2026. Three private U.S. companies — OpenAI, Anthropic and Google — now control the most advanced AI systems on Earth, and every government at that G7 table runs critical infrastructure on their models. No elected parliament built these systems. No international treaty governs them. They showed up to Évian not as vendors, but as peers to heads of state, as Memeburn's reporting on the summit made plain.

The backdrop to the lunch made the stakes visceral. Just days before the summit, the U.S. government had applied export controls to Anthropic's most capable model — the first time Washington had ever weaponised an AI model as a direct instrument of foreign policy. Amodei arrived at the table with his best product effectively switched off for international users, his company deadlocked with the Trump administration, and his largest investor, Amazon, having reportedly triggered the action. According to analysts covering the session, he needed a diplomatic resolution or Anthropic's frontier position would erode while competitors continued shipping.

The three men at that table embody distinct and consequential philosophies about what AI is for and how it should be governed.

Dario Amodei, who co-founded Anthropic after leaving OpenAI, has built his company around the proposition that safety and capability are not in conflict. He has argued publicly that human-level AI could arrive as early as 2026 or 2027, describing rapid progress toward automating complex software engineering as a near-term reality. At the G7, according to CNBC's reporting, he named his two priorities for international cooperation with precision: structured access to frontier models, and a chip trade framework that excludes China. His position is that the most dangerous thing the democratic world could do is allow frontier AI to develop without allied coordination.

Demis Hassabis, the neuroscientist-turned-entrepreneur who built DeepMind before its acquisition by Google and now leads the merged Google DeepMind, has consistently emphasised what he calls the scientific mission of AI. At the G7, he pushed for what the AI Chronicle described as 'scientific diplomacy' — the argument that AI can solve civilisational problems from climate change to drug discovery, but only within a stable framework of public-private cooperation. Hassabis has been more cautious than Amodei on timelines, arguing that today's systems, however impressive, remain some distance from the full range of human cognitive capability. His strategic bet is that Google's integrated stack — from custom chips to Gemini models to cloud infrastructure — gives it a durable advantage that pure-play AI labs cannot replicate.

Sam Altman occupies the most exposed position of the three. OpenAI's S-1 filing for what analysts describe as a potential trillion-dollar IPO was in process at the time of the Évian summit, meaning regulatory clarity was not merely a geopolitical preference but a direct financial necessity. At the G7, Altman called for an international forum to establish globally accepted standards for testing AI, provide expert and impartial analysis of capabilities and risks, and serve as a venue for cooperation among nations, according to an OpenAI briefing cited by Memeburn. That is a vision of an institutionalised global AI order — with OpenAI positioned near its centre.

Below the summit theatre, the economics of the AI industry are becoming harder to read. A PwC study published in April found that nearly 74 percent of AI's economic value is being captured by just 20 percent of organisations, revealing what the firm described as a stark and widening divide between a small group of AI leaders and the majority of businesses still stuck in pilot mode. The research showed that top-performing companies are not simply deploying more tools — they are using AI as a catalyst for growth and business reinvention.

The BCG AI Radar, published in June, found that nearly three-quarters of CEOs have made themselves their company's chief decision-maker on AI — twice the share of a year ago — and that corporations plan to roughly double their AI spending as a share of revenues in 2026. Four out of five CEOs told BCG's researchers they are more optimistic about AI returns than they were twelve months ago. And yet 60 percent of CEOs surveyed by BCG admitted they had deliberately slowed AI implementation due to concerns about errors and malfunctions, according to the World Economic Forum's reporting on the findings.

The capital markets are rendering their own verdict in real time. Forbes reported in April that Alphabet surged on first-quarter earnings while Microsoft and Meta faced investor caution despite beating revenue estimates — because Alphabet demonstrated demonstrable AI revenue through Google Cloud's 63 percent growth to $20 billion, while its peers could not yet convert ballooning capital expenditure into equivalent returns. Sundar Pichai told investors that AI solutions were the primary growth driver for Google Cloud for the first time. Meanwhile, Amazon's full-year capital expenditure budget is forecast to reach $200 billion — 53 percent above the prior year — even as free cash flow fell sharply in the first quarter.

Macron captured the anxiety of every government at Évian in a single warning, as reported by Memeburn: nobody will buy American AI if they fear it can be switched off at any moment. That sentence is not just a diplomatic rebuke — it is a statement of the fundamental structural tension in the emerging AI order. The Fable 5 export control episode proved that frontier AI is now a geopolitical instrument, not merely a commercial product. And the man who built Fable 5 — Amodei — was sitting across the table from the officials who shut it down.

What comes next will be determined less by who builds the most powerful model and more by who writes the rules that govern access to it. The three architects of intelligence who gathered in France understand that better than anyone. The question is whether the governments at their table are ready to move as fast as the technology they are trying to govern.