Instead of an outright refusal, Fable 5's safety classifiers silently switch sensitive queries about cybersecurity or biology to the less-capable Opus 4.8 model. This layered approach maintains functionality while containing perceived risks, though it can lead to user confusion when performance unexpectedly drops for certain prompts.
Just before the launch, Anthropic publicly warned about the dangers of frontier AI, urging for a "brake pedal." This cautious positioning, possibly intended to demonstrate responsibility, may have been taken at face value by the government, contributing to the rapid and decisive shutdown of their most advanced model.
To pull Fable 5, the government used an export control order. While aimed at preventing foreign access, its broad application to all foreign nationals—including a company's own US-based employees—made a selective block impractical. This forced a complete worldwide shutdown for all users, including domestic ones.
Fable 5’s key advantage isn't marginal improvements on simple queries. Its performance lead grows significantly with task length and complexity. This indicates a shift toward models built for sustained, long-form work like codebase migrations or complex research, representing a new tier of AI capability.
While seemingly a business setback, having its model banned by the U.S. government for being too powerful sends a strong market signal. This event implies Anthropic's technology is ahead of competitors like OpenAI and Google, potentially boosting its brand reputation and valuation ahead of its IPO.