Summary of Key Points
After being banned by the U.S. Department of Commerce for over half a month, Anthropic's cutting-edge AI model Fable 5 has finally been released, but its usability has significantly diminished. Ordinary users find it difficult to access all its features even after paying. To comply with regulations, Anthropic has implemented stricter security measures, which has led to widespread criticism from the industry regarding the arbitrary nature of U.S. government regulation. This incident has also sparked a shift in preference towards open-source models, with many small and medium-sized businesses opting for the lower-risk alternatives.
1. Ordinary Users: No More Advanced Experience for Money
With the release of Fable 5, users have realized that the days of being able to use the most advanced model simply by paying are over. Anthropic has stipulated that, until July 7th, paid users can only use half of their allocated usage quota on Fable 5 each week. If a “security check” is triggered, they will automatically be redirected to the older Opus 4.8 model. For example, Ping Fuqiang, a Stanford economics PhD and a loyal Anthropic user, was so angered by the restrictions that he cursed Anthropic’s CEO, Amodi. Another programmer spent $321 on Fable 5 for coding purposes but found that three-quarters of his tasks were redirected to the older model. Even more frustratingly, users discovered in system logs that the reason given was that they were “too stupid to use Fable”—even Anthropic’s own employees admitted they didn’t expect users to check the logs.
Despite Fable 5’s impressive performance in certain benchmarks (such as ranking first in the “Remote Labor Index” and outperforming GPT-5.5 in physical animation tasks), ordinary users cannot benefit from these capabilities. This has dampened their enthusiasm for spending money on the model, leading them to wonder, “If I can’t use the latest features even after paying, why not just use the free version?”
2. Anthropic: Imposing Strict Restrictions to Gain Releasing Permission
To get Fable 5 back online, Anthropic made several concessions:
1. Strengthened Security Checks: The newly released Fable 5 includes a more stringent “classifier” system that may block even harmless programming questions. Anthropic claims this is necessary to ensure safety and that users must phrase their queries in a very clear manner to avoid being blocked.
2. Appealing for Fairness + Demonstrating Commitment: In an official blog post, Anthropic argued that other models less powerful than Fable 5 also have the same vulnerabilities. They also claimed that the new classifier can block 99% of “dangerous requests,” with the remaining 1% being unlikely to pose a threat.
3. Promoting Industry Standards: Anthropic is working with partners like Amazon to establish clear industry guidelines to prevent similar bans in the future, citing the lack of clear standards for AI models to bypass security restrictions.
In essence, Anthropic is saying, “We have already implemented very strict controls; the authorities and our competitors should stop causing problems.”
3. Industry Criticism: Arbitrary U.S. Government Regulation
The industry’s main dissatisfaction with the Fable 5 ban and subsequent release lies in the arbitrary nature of U.S. government regulation:
- Lack of Transparency: Mooreck, a professor at the Wharton School of Business, criticized the lack of clear explanations from the authorities regarding the reasons for the ban and the security assessment process, questioning its credibility.
- Unclear Procedures: Matt Peir, former head of technology at the White House, pointed out that there are no established rules, leaving companies unsure of what to do. For example, the release of GPT-5.6 was delayed due to government pressure. Who knows which model will be the next target?
- High Risks for Small Businesses: Small and medium-sized businesses are particularly worried. For instance, a legal AI company named Legion sued the government because some of its employees were unable to use Fable 5. With open-source models being the foundation of their products, they fear that bans could be arbitrary and destructive.
The industry unanimously calls for clear, transparent regulations from the authorities to end this chaotic situation.
4. Industry Trends: The Rise of Open-Source Models
The Fable 5 incident has highlighted the risks associated with closed-source models (such as those from Anthropic and OpenAI, which do not share their code). Open-source models, being more flexible and less risky, are becoming increasingly popular:
- Shift to Open-Source Models: Many entrepreneurs are turning to open-source models for their products, as they don’t have to worry about them being banned by the authorities. According to predictions, the usage of open-source models in China could double by 2026, while the share of closed-source models in the U.S. could be halved.
- A Repeat of History: Similar to how open-source Linux overtook closed-source Windows, open-source models may emerge victorious in the face of restrictive regulations. Open-source models offer greater flexibility and lower risks.
This incident could potentially revolutionize the AI industry.
Conclusion
The release of Fable 5 marks not the end but a new beginning. User interest in paid models has waned, Anthropic has had to compromise for regulatory compliance, and the issues with government regulation have become evident. The advantages of open-source models are becoming increasingly clear. The rules of the AI industry are quietly changing.