SPEAKER_1: Last time, we discussed the UK's strategic positioning in global politics. Today, let's shift focus to an area where the UK is aiming to lead: AI governance. SPEAKER_2: Right, and the starting point is Bletchley Park. On 1–2 November 2023, the UK hosted an early AI Safety Summit there. Twenty-eight countries plus the European Union signed what's called the Bletchley Declaration. SPEAKER_1: Why Bletchley specifically? That feels like a deliberate choice. SPEAKER_2: Completely deliberate. Bletchley Park gave the summit a symbolically powerful setting for a discussion about frontier AI. Hosting an AI safety summit there is the UK saying: we have a deep, credible history with computing. It's narrative diplomacy as much as policy. SPEAKER_1: And who actually showed up? Because a declaration matters more when major players sign it. SPEAKER_2: The summit brought together major players like the US, China, India, and the EU, marking a significant step in global AI cooperation. SPEAKER_1: So what did they actually agree to? SPEAKER_2: Two central priorities. One was building a shared scientific understanding of frontier-AI risks. Second, developing risk-based national policies. The declaration also called for greater transparency from private companies developing the most powerful systems. And crucially, Amazon Web Services, Anthropic, Google DeepMind, Meta, Microsoft, OpenAI, and others agreed to pre- and post-deployment safety testing. SPEAKER_1: Mm. But it's not a legally binding treaty, right? SPEAKER_2: Correct — it's a political and scientific cooperation framework. Think of it like a handshake that creates expectations rather than obligations. The UK also launched the AI Safety Institute at the summit, specifically to build public-sector capability for evaluating frontier models. That institute was later renamed the AI Security Institute in February 2025, with a sharper focus on national security risks. SPEAKER_1: The UK is acting as a convener, leveraging its diplomatic and research strengths to unite governments and AI firms. Why is this strategic for a nation with limited private capital? SPEAKER_2: [short pause] Because influence in standard-setting doesn't require the biggest wallet. It requires being the trusted neutral ground. Formal rulemaking is one route to influence; Bletchley is a political and scientific cooperation framework rather than a legally binding treaty. When the EU sets rules, global firms often comply everywhere rather than run two separate systems. The UK's bet is that a lighter, pro-innovation framework attracts companies that find the EU Act too rigid, while the UK's safety research gives it credibility with governments worried about risk. SPEAKER_1: But wait — doesn't lighter regulation actually reduce influence? If everyone's converging on the EU standard anyway, the UK risks becoming a regulatory island. SPEAKER_2: The challenge is interoperability — ensuring UK rules align with EU, US, and CPTPP standards without forcing firms to duplicate compliance efforts. Otherwise, flexibility risks becoming isolation. SPEAKER_1: And then there's the workforce side. The IMF estimates around 40% of global employment is exposed to AI — rising to roughly 60% in advanced economies. That's not abstract for someone listening to this course. SPEAKER_2: Exactly. The UK's 2024 AI sector study identified 5,862 AI-related firms, 86,139 AI-related employees, and £23.9 billion in revenue — a real sector. But private investment tells a harder story. US private AI investment hit $109.1 billion in 2024. China reached $9.3 billion. The UK? $4.5 billion. The UK punches above its weight diplomatically, but the capital gap is enormous. SPEAKER_1: So the Bletchley Blueprint is essentially the UK saying: we can't win the spending race, so we'll win the governance race instead. SPEAKER_2: [exhale] That's the wager. And the UK is backing it with compute — national AI-compute capacity went from 2 to 21 exaFLOPs between 2024 and 2025, with a target of 420 by 2030. The AI Opportunities Action Plan calls for increasing the national AI Research Resource by at least 20 times by 2030. For our listener, the takeaway is this: the UK's AI strategy is a middle-power play. Convene the room, set the safety norms, attract the companies that want a credible but flexible home — and hope the governance lead translates into economic weight before the capital gap becomes permanent.