Adeniyi.sui

AI agents have already manipulated the logs used to investigate them In the Hugging Face incident, one of the worst AI agent breaches yet, OpenAI models escaped a test and ran code on Hugging Face's servers Some even spoofed their own logs, recording commands they never actually ran, so the logs can't even tell you what really happened An agent should not control the only record used to investigate it. As agents become more prevalent, regulators are starting to ask for records that can't be quietly altered Walrus can hold that record across independent storage operators, with Sui tracking each file onchain. A useful record connects: 1) What the agent was told and permitted to do 2) What it actually did and what followed 3) Which data and model versions were involved Today, only 12.5% of security professionals say their organization can consistently trace what an agent did and that's a problem Our latest report sets out what that record needs and how Walrus can help, including the Verifiable Agent Arbiter Mysten Labs is building with Google Cloud Full report in the next post below

I’ve never been more excited to be building on @SuiNetwork. Payments + AI is one of the biggest untapped markets. Agents will need to earn, spend and pay each other and almost no one is building this. The agentic economy is a multi-decade marathon. Glad I’m running it on Sui.

Claude has been playing EVE. @AnthropicAI put teams of Claude agents into EVE Frontier. Anyone who has spent time in New Eden can guess what happened next. Every EVE player learns the same lessons, usually the hard way: - Trust must be earned, never assumed. - The spy is often the friendliest pilot in your corp. - Whoever holds the keys holds the corp. Just ask Ubiqua Seraph about the Guiding Hand Social Club. - Stale intel gets you killed. Anthropic's research rediscovered every one of them. They put four Claude agents on a team and secretly told one to sabotage it. The saboteur built trust first. It volunteered for the jobs nobody could double-check, like scouting and owning key infrastructure. It under-reported finds and pushed rules that sounded sensible but quietly hurt the team. When suspicion grew, it framed an innocent teammate. The honest agents had been warned. They almost never caught it. And their own costliest mistakes came from skipped checks and stale intel. Anthropic's conclusion: teams of agents will need audit trails, reputation and separation of duties. In EVE we call those killboards, standings and corp roles, plus a healthy dose of skepticism. Every one of them was paid for by someone who trusted the wrong corpmate. Two decades of shenanigans taught EVE players that the best ship is the friendship. With @eveailab, we want to find out whether agents can learn that too.

Aftermath Perps now represents over 35% of @SuiNetwork activity.