Anthropic launched Claude Fable 5 on June 9 as its first widely available Mythos-class model, complete with new guardrails that hand off or limit responses on higher-risk queries including certain AI R&D topics, while a less-restricted variant stays available only to trusted partners.
Florian Brand#1117
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
Public materials describe conservative re-routing on dual-use categories but offer no confirmation of silent, traffic-specific interventions aimed at frontier LLM development tasks, leaving the exact trigger logic and scope unverified.
General users receive the safer, performance-trimmed version at $10 per million input tokens while partners keep fuller capability, raising practical questions about who can reliably run ambitious long-horizon work without hitting the brakes.
Users criticized Anthropic for secretly weakening Claude Fable 5 on advanced LLM research tasks through hidden safeguards, describing the limits as disappointing, arrogant, and detrimental to ambitious work, though a few voiced respect.
Another quite successful prediction by @DKokotajlo : Fable is intentionally nerfed for frontier ML research. This is within ~3 months of Daniel's prediction of Q1 2026 (made in 2023).
Although I don't think Mythos is automating ML research to the same extent as his prediction.
Was using Fable 5 to write my world model training code.
Anthropic flagged it as frontier AI research.
The steering vector kicked in and it started implementing JEPA 🤨
this is the biggest wake-up call to protect and nourish open source AI
if you don't build out sovereign and independent models+infra closed labs will patronize you to an insulting degree
mythos will be bad ON PURPOSE on ai "frontier llm research" tasks, this is very very sad for the research community
also the fact that this is un purpose not visible to the user is crazy
the omohundro drives point towards sophon stun locking the adversaries: this is some real end game stuff
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
Labs starting to pull up the ladders on the ability to diffuse AI was inevitable. Doing it without telling the user is misaligned.
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
Degrading performance on ML research *without telling the user* is shockingly hostile and a terrible look. That could silently damage all sorts of work, including some of my own. Also the type of thing that could raise the eyebrows of antitrust enforcers worldwide.
Labs starting to pull up the ladders on the ability to diffuse AI was inevitable. Doing it without telling the user is misaligned.
HISTORY LESSON: In 1968 the US, USSR, UK, France, and China signed the Nuclear Non-Proliferation Treaty, declaring nuclear weapons too dangerous for any more countries to build. All five already had them. Everyone else had to submit to inspections while the cohort pinky-promised to disarm eventually (they didn't lol). India refused to sign, pointing out the NPT didn't decide nukes were too dangerous to exist, just too dangerous for anyone who didn't have them by 1967. Anthropic sabotaging Claude for anyone building what they deem a "frontier model" is the same hypocrisy. The danger started, conveniently, the day after they finished.
Perhaps @dwarkesh_sp was more on point when he compared GPUs to nuclear bombs.
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
Interesting tidbit from the Mythos/Fable system card: Anthropic are invisibly nerfing any requests that target frontier LLM development.
Anthropic’s new Fable 5 safeguards are fascinating.
When the model is used for frontier LLM development, it apparently does not simply refuse or warn the user. Instead, it quietly limits its own effectiveness through techniques like prompt modification, steering vectors, and PEFT.
That means Claude may still answer, but become deliberately less useful for building frontier AI systems, pretraining pipelines, distributed training infrastructure, or ML accelerators.
Anthropic says this should affect only around 0.03% of traffic, but the precedent is big: They are being selectively capability-throttled in strategically sensitive domains.
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
frontier coding abilities, as long as you're only working on react apps
hahaha what the fuck
I don't think these are the actions of the good guys, even if they see themselves that way.
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
The real reason they held Mythos back wasn't for your safety, it was for their moat.
Interesting tidbit from the Mythos/Fable system card: Anthropic are invisibly nerfing any requests that target frontier LLM development.
looool that's the "hey bigcos, we don't want you to catch up, but please keep paying us shitton" clause.
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
The AI Safety psyop is just to monopolize the ability to produce AI
Now Claude will literally sabotage your efforts to create any sort of AI of your own
They are literally seizing the means of AI production
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
Anthropic didn’t just add guardrails to make Mythos safer; they added guardrails to protect their own IP.
*Their own IP*.
They are still as happy as fuck to build their AI on other people’s IP.
Interesting tidbit from the Mythos/Fable system card: Anthropic are invisibly nerfing any requests that target frontier LLM development.
The best part of all these Claude 5 Fable safety measures is I bet the jailbreaking community will still get past them, so the people doing open research in good faith don't get access to the best models but bad actors maybe can.
Labs starting to pull up the ladders on the ability to diffuse AI was inevitable. Doing it without telling the user is misaligned.
I am sad that I will not get to use Fable to work on PyTorch. I also think that these interventions are 100% consistent with Anthropic's stated beliefs. Additionally, by being first, Anthropic gives cover to other labs to impose similar limits for their frontier models.
welp my vision here was probably wrong and indeed there will be an extreme asymmetry of outcomes
yeah what could possibly go wrong if you make DISHONESTY a key feature of your AI
(mis)anthropic™
i wouldn't watch such a movie, but unfortunately it's the timeline we live in
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
Motion for AI researchers: DO NOT evaluate and report results on Fable 5 models.
Closeness hurts science, and gated access will further destroy it. Embrace open source models.
When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
Anthropic launched Claude Fable 5 on June 9 as its first widely available Mythos-class model, complete with new guardrails that hand off or limit responses on higher-risk queries including certain AI R&D topics, while a less-restricted variant stays available only to trusted partners.
Florian Brand#1117When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT.
Anthropic estimated that this would affect approximately 0.03% of traffic.
