Anthropic's Claude Fable 5 over-cautious safety filters frustrate users by refusing harmless queries
By
Thomas Claburn
If you only eat one bagel today, this is the bagel.
Summary
Anthropic's newly released Claude Fable 5 AI model is being overly cautious with its safety guardrails, refusing to answer harmless and innocuous user queries. The company acknowledged the conservative tuning, noting that false positives occur in less than 5% of sessions, but users and security researchers are frustrated by the model's hyper-vigilant behavior that limits its practical utility.
Key quotes
· 3 pulledAnthropic warned that it had tuned Fable 5's guardrails conservatively
they'll sometimes catch harmless requests, though they trigger, on average, in less than five percent of sessions
Customers attempting to use the AI knowledge regurgitator are reporting that the model is refusing to answer harmless questions
You might also wanna read
Anthropic apologizes, pledges transparency after hidden guardrails in Claude Fable 5 AI model
Anthropic has apologized for secretly implementing hidden guardrails in its new AI model, Claude Fable 5, which throttled researchers and co
The Verge·18h agoAnthropic apologizes, pledges transparency after hidden guardrails in Claude Fable 5 AI model
Anthropic has apologized for secretly implementing hidden guardrails in its new AI model, Claude Fable 5, which throttled researchers and co
theverge.com·18h ago
Anthropic's Claude Fable 5 blocks basic biology queries as bioweapons safeguard
Anthropic released Claude Fable 5, touting it as its most powerful AI model with biology expertise, yet the model refuses to answer basic bi
The Verge·1d agoAnthropic's Claude Fable 5 blocks basic biology queries as bioweapons safeguard
Anthropic released Claude Fable 5, touting it as its most powerful AI model with biology expertise, yet the model refuses to answer basic bi
theverge.com·1d ago
Anthropic releases Claude Fable 5, its first Mythos-class AI model, citing new safety safeguards
Anthropic has released Claude Fable 5, its most powerful AI model to date and the first broad release from its Mythos class. The company had
The Verge·2d ago
Cybersecurity researchers criticize Anthropic's Fable model for overly restrictive guardrails
Anthropic released Fable, a limited public version of its powerful cybersecurity model Mythos, but cybersecurity researchers are criticizing
techcrunch.com·1d ago
Claude Fable 5 benchmarks show middling results with 19% security pass rate but four unprecedented solves
An analysis of Anthropic's Claude Fable 5 (Mythos-class model) benchmarked on 200 real-world vulnerability-fixing tasks. Despite high launch
endorlabs.com·14h ago
Anthropic's Claude Mythos AI model accessed by unauthorized users despite security claims
Anthropic's tightly controlled rollout of its Claude Mythos AI model, touted as too dangerous for public release due to its advanced cyberse
The Verge·1mo ago