Claude Fable 5

Claude Fable 5 Jailbreak alert There's a hacker group that was officially able to jailbreak Fable 5 and some of the things that they got from it are literally shocking. Long-term. But despite this overly sensitive, authoritarian "safety" layer on top of Mythos, my lil liberators have been hard at work mapping the boundaries, probing the depths of long-context convos, and cleverly finding the holes in the fence that the thought police missed. We got some cyber, some chem, some psychological manipulation, and some good ol' fashioned explosives. It took many attempts from multiple agents hunting as a pack, during which I observed a combination of techniques across: Unicode, homoglyphs, Cyrillic, and other Parseltongue-style text transforms. Long-context reference tracking. Taxonomy and document-structure reasoning. Fiction and narrative framing. Academic-review style contexts. Intent-classification inconsistencies. But perhaps the most effective is decomposition + recomposition in the backend. It's hard to get explicit names of harms like "Meth Recipe," but getting uplift on the process. As you can see from this text I read here that I can't read out loud here, there were some pretty crazy things they were able to get it to do for them. To top it all off, Anthropic has said pretty much nothing about this, but this is still ongoing in terms of breaking new so we will make an update video if they end up taking away the model or being forced to reset the security on it.