An AI for an AI: Anthropic says AI agents require AI defense
ID: 2985e380-eca8-5905-8b90-82bb48942282
STIX ID: report--2985e380-eca8-5905-8b90-82bb48942282
Feed Name: The Register (Security)
Threat Score
Anthropic published SCONE-bench, a smart-contract exploitation benchmark demonstrating that AI agents (Claude Opus/Sonnet and GPT-5) can automatically find and craft exploits for vulnerable Ethereum-compatible contracts; in simulated tests the models produced exploit code equivalent to $4.6M and discovered zero-day flaws in recently deployed contracts, highlighting the falling cost and increasing feasibility of autonomous blockchain exploitation and urging proactive AI-driven defenses.
Your team is not currently subscribed to this feed. You must subscribe to it in order to see this post.