EN ▾
SNAPE REP RT Front page →
ConfirmedAI safety04 Nov, 08:27 UTC

Meta has paid about $1.2m in bug bounties for its Muse AI agent, including $300,000 for one flaw

SecurityWeek illustration for its report on Meta's Muse bug bounty
SecurityWeek

A pre-launch flaw exposed conversations and tokens on staff test machines; $90,000 went to a prompt-injection page that faked browsing-tool messages. Of 630+ reports, 64 were valid.

Read at SecurityWeek Share

Sources

More on AI safety
More from Snape ReportFront page →