Anthropic Locks Mythos After Corporate Network Attack Success

Failed to add items

Sorry, we are unable to add the item because your shopping basket is already at capacity.

Add to cart failed.

Please try again later

Add to wishlist failed.

Please try again later

Remove from wishlist failed.

Please try again later

Follow podcast failed

Unfollow podcast failed

Anthropic Locks Mythos After Corporate Network Attack Success

Listen for free

View show details

Anthropic just released Claude Opus 4.7, and Adobe and Figma's stocks immediately dropped over 2 percent because this thing can apparently handle your hardest coding and design work without supervision—oh, and it can check its own homework before reporting back. The wild part is they're keeping an even scarier version called Claude Mythos locked down because it completed 3 out of 10 attempts at a 32-step corporate network attack that takes human hackers 20 hours, and they're only giving access to Amazon and Microsoft. Meanwhile they've updated the tokenizer so you're using up to 35 percent more tokens for the same work at the same per-token price, which is technically not a price increase but functionally absolutely a price increase.

No reviews yet