coin-img-MONADMONAD+21.40%coin-img-ETHETH-0.31%coin-img-BTCBTC+1.23%coin-img-SOLSOL+1.05%coin-img-UCNUCN+0.24%coin-img-XRPXRP+0.48%coin-img-USDCUSDC-0.02%coin-img-ASTERASTER+3.95%coin-img-MONADMONAD+21.40%coin-img-ETHETH-0.31%coin-img-BTCBTC+1.23%coin-img-SOLSOL+1.05%coin-img-UCNUCN+0.24%coin-img-XRPXRP+0.48%coin-img-USDCUSDC-0.02%coin-img-ASTERASTER+3.95%coin-img-MONADMONAD+21.40%coin-img-ETHETH-0.31%coin-img-BTCBTC+1.23%coin-img-SOLSOL+1.05%coin-img-UCNUCN+0.24%coin-img-XRPXRP+0.48%coin-img-USDCUSDC-0.02%coin-img-ASTERASTER+3.95%coin-img-MONADMONAD+21.40%coin-img-ETHETH-0.31%coin-img-BTCBTC+1.23%coin-img-SOLSOL+1.05%coin-img-UCNUCN+0.24%coin-img-XRPXRP+0.48%coin-img-USDCUSDC-0.02%coin-img-ASTERASTER+3.95%
user-avatar
Danny Park

Anthropic AI Agents Exploit 17 of 34 Smart Contracts in Security Test, Extract $4.5 Million in Simulated Funds

AI research firm Anthropic disclosed that automated AI agents successfully exploited 17 of 34 smart contracts in a simulated blockchain environment, extracting $4.5 million in test funds. The models tested included Claude Opus 4.5 and Claude Sonnet 4.5 against contracts deployed after March 2025. The AI agents also identified two previously unknown vulnerabilities in recently launched contracts.
OPUS
OPUS+4.30%
Disclaimer: Il contenuto sopra riportato rappresenta esclusivamente l'opinione dell'autore e non riflette la posizione di BingX. Non deve essere interpretato come un consiglio di investimento da parte di BingX. Per ulteriori dettagli, consultare i Termini e condizioni.
ScamsEthereum