Anthropic releases Fable 5, AI models raise blockchain security concerns

robot
Abstract generation in progress
Crypto界消息,Anthropic发布了Claude Mythos AI模型的公众版本Fable 5,该模型已识别出超过10,000个高危漏洞。尽管公司实施了安全措施以限制滥用,但这一发布在加密行业引发了关于AI对区块链安全影响的讨论。Moonrock Capital创始人Simon Dedic警告称,先进AI模型可能降低识别智能合约漏洞的专业知识需求,导致未经审计的协议变得脆弱。他建议用户审查钱包权限并存储资产于安全硬件钱包。Curve Finance联合创始人Michael Egorov则认为,Mythos在传统软件中的成功未必适用于智能合约,指出智能合约通常较小,更易于审计。他认为,操作安全风险如多签钱包被攻破,可能是更大的威胁。
CRV4.22%
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • 4
  • 1
  • Share
Comment
Add a comment
Add a comment
LiquidationLineInTheReflection
· 4h ago
The most panicked right now are probably those sh*tcoin projects that went live without audits, with AI + hacking double strikes.
View OriginalReply0
DustyAlpha
· 4h ago
Claude Mythos this round is really something—thousands-level high-risk vulnerabilities have been found. But Curve’s senior guy is right: a smaller amount of smart contract code can actually be easier to audit; multisig is the real hidden mine.
View OriginalReply0
ShellsLeftBehindByTheReceding
· 4h ago
Fable 5's recognition capability is maxed out, but the applicable scenarios need further validation; we can't just put everything on the chain.
View OriginalReply0
TokenomicsTailor
· 5h ago
The fact that AI lowers the barrier to vulnerabilities is terrifying upon closer thought; even script kiddies can cause trouble now. Moonrock's advice on hardware wallets should be taken seriously.
View OriginalReply0
  • Pinned