The people building actual frontier AI are gathering in London on June 12th, and the speaker list tells you exactly where the technical edge is moving.

The Summary

  • RAAIS 2026 announces headline speakers for its 10th annual Research and Applied AI Summit, featuring teams from Anthropic, Google DeepMind, and orbital data center pioneers
  • Speaker themes cluster around open-ended agents, medical AI, world models, and space-based compute infrastructure
  • This isn't a product launch event. It's where research directors share what's actually working in production

The Signal

RAAIS has been running for a decade, which means it predates the ChatGPT moment by seven years. The conferences that survive that long do so because they stay close to actual research, not hype cycles. This year's lineup reflects where serious AI labs are placing their bets.

Anthropic's research team is presenting on open-ended agents. Not chatbots. Not copilots. Agents that pursue goals without constant human handholding. Google DeepMind is covering world models, the technical foundation for agents that can simulate consequences before acting. These aren't separate trends. They're two sides of the same capability: AI systems that model reality well enough to act autonomously within it.

"The research presenting at RAAIS shows the infrastructure layer for Web4 is being built right now, in production environments, not labs."

The medical AI track matters because healthcare is where agent capabilities meet regulatory reality first. If your AI agent is going to make decisions that affect human bodies, you need interpretability, safety guarantees, and audit trails that don't exist yet. The teams presenting here are building them.

Then there's the orbital data center piece. Space-based compute isn't science fiction anymore. It's a response to three hard limits: energy availability, cooling costs, and latency to edge devices. When your agent infrastructure needs to serve 8 billion people running personal AI, you can't rely on terrestrial data centers alone.

The Implication

If you're building agent companies or trying to figure out where the infrastructure bottlenecks are, this speaker list is a roadmap. Watch for published talks after the event. The gap between what gets presented at RAAIS in June and what ships as products by December is getting shorter every year.

For everyone else: the future of work isn't about AI replacing humans. It's about AI agents doing work humans never had time for in the first place. The technical foundations being discussed at this conference determine how fast that future arrives.

Sources

State of AI