AI TRENDS | David Sacks Says Anthropic’s Fable Model Raised Security Concerns Disputed by U.S. Government

David Sacks, co-chair of the U.S. President’s Council of Advisors on Science and Technology, said a security dispute involving Anthropic’s newly released model “Fable” has prompted regulatory attention. According to Odaily, Sacks said he has communicated with multiple parties about the situation and described the core issue as security concerns tied to Fable, which he characterized as a commercial version of a Mythos-class model. Sacks said Anthropic publicly described the vulnerability as “not serious,” but said the U.S. government and testing parties disagreed, arguing it was sufficient to affect the model’s security and could involve risks related to “cyber weapon operability.” He also criticized Anthropic for previously emphasizing “safety first” but, in this case, prioritizing keeping the consumer version online rather than addressing the security issue first. Sacks said the matter should not be conflated with other prior defense or regulatory disputes. He added that the U.S. government still recognizes Anthropic’s technical capabilities and said the issue could have been resolved quickly, with the initiative resting with Anthropic.