{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreih7ogfpdkeumx5n4t2tlyxj6op5gl5mbkop3p5sttc7rxwkxaalze",
"uri": "at://did:plc:ajcrkmnlj6rxdk7rltijv227/app.bsky.feed.post/3mkayfbjr3vy2"
},
"coverImage": {
"$type": "blob",
"ref": {
"$link": "bafkreicgahbonucumx2yjvi66kv7zfngxm3udrk35hhnmazdlp2kdus5wu"
},
"mimeType": "image/jpeg",
"size": 75326
},
"path": "/tech-industry/cyber-security/how-a-cavalcade-of-blunders-gave-unauthorized-users-access-to-claude-mythos-restricted-model-accessed-by-third-parties-thanks-to-knowledge-from-data-breach",
"publishedAt": "2026-04-24T15:12:08.000Z",
"site": "https://www.tomshardware.com",
"tags": [
"Cybersecurity",
"Tech Industry"
],
"textContent": "Unauthorized individuals have accessed Anthropic's new Mythos cybersecurity-focused AI model, despite the developer locking it down to just a handful of companies. Considering the AI was purposefully designed to find zero-day exploits and offer viable fixes, the breach raises questions about Anthropic's own security, and why Mythos couldn't protect it.",
"title": "How a cavalcade of blunders gave unauthorized users access to Claude Mythos — restricted model accessed by third parties, thanks to knowledge from data breach"
}