Designing for agents that make mistakes. Agents will get things wrong. Diego Socolinsky shares how architects are planning for it, and where recovery fits when they do.
The AI supply chain. Your AI stack is only as trustworthy as everything plugged into it. Find out where attackers are likely to look next.
Resilience for the agentic era. What to fund, test, and change in 2027 so you can recover when AI moves faster than oversight.
Agents have identities now. What's next? The tools have arrived; the habits haven't. David Gildea on what changes in 2027, and why protecting and recovering identity with context matters more than ever.
2027's biggest agentic AI surprises: one good, one bad. Each speaker makes one call on each side, and they won't necessarily agree.