Anthropic scrambles after Trump administration freezes its top AI models amid security concerns over potential 'jailbreak' vulnerabilities.

The Trump administration's freeze on Anthropic's AI models has blindsided the industry, raising questions about Washington's approach to technology. The rapid response has sparked concerns that it could halt new model deployments for all frontier model providers.

Sources:
Bloomberg.comFinancial Times
Trending Today
Tab background
Sources: Bloomberg.comFinancial Times
The Trump administration's recent order to freeze Anthropic's top AI models, including Fable and Mythos, has sent shockwaves through the tech industry. This unprecedented move, which bars foreign access to these advanced systems, was prompted by concerns over potential 'jailbreak' vulnerabilities that could compromise security.

Anthropic was given 90 minutes to comply with the directive, which was issued without detailed explanations. The company has since mobilized a senior technical team to Washington to address the fallout from the decision. According to sources, the order followed a report from Amazon researchers who identified a potential way to bypass the guardrails of the Fable model, raising alarms about the extraction of sensitive information.

The Department of Commerce has implemented export controls on Anthropic's models, effectively halting their use by foreign nationals. This swift action has raised questions about the administration's approach to AI technology and its implications for the industry. Notably, the freeze does not extend to competing models from OpenAI, despite similar capabilities being present.

Anthropic has contested the government's targeting of its models, arguing that the vulnerabilities identified are not unique to its technology. The Pentagon has also flagged Anthropic as a supply-chain risk to national security, further complicating the situation as the company navigates litigation over this designation.
Sources: Financial TimesBloomberg.com
The Trump administration has ordered Anthropic to freeze its advanced AI models due to security concerns over potential 'jailbreak' vulnerabilities, raising alarms in the tech industry. The directive, issued without detailed explanation, has blindsided the company and prompted immediate compliance efforts.
Section 1 background
The Headline

Anthropic's AI models frozen by US government

Key Facts
  • The Trump administration's decision to force Anthropic to suspend its most advanced AI models, days after the $900bn company released them to the public, has blindsided the AI industry and raised questions over Washington's approach to the technology.Financial Times
  • The Department of Commerce on Friday imposed export controls on Anthropic's newest models, Fable and Mythos, barring foreign nationals from using the technology and prompting the company to suspend both systems for all users.Financial Times
  • Anthropic was given 90 minutes to comply and was not provided with detailed concerns before the order was issued, according to a person close to the company.Financial Times
  • Over the weekend, Anthropic has flown in a senior technical team to Washington in an effort to contain the fallout.Financial Times
  • The order was issued after it was discovered that the Fable 5 AI model could be “jailbroken,” or bypass its guardrails.Bloomberg.com
  • The rapid response from the US government has raised concerns that it could “halt all new model deployments for all frontier model providers.”Bloomberg.com

Related Videos

Trump administration forces Anthropic to pull models worldwide
Trump administrationAnthropicAI modelsartificial intelligencetechnology news
‹
›
Section 2 background
Background Context

Context on Anthropic's AI model freeze

Key Facts
  • The directive followed a report by Amazon researchers identifying a potential “jailbreak” in Fable — a way of bypassing the model's guardrails and extracting information about security vulnerabilities in software.Financial Times
  • Amazon first shared its findings with Anthropic.Financial Times
  • Andy Jassy, Amazon’s chief executive, discussed the issue with US officials on Friday.Financial Times
  • Anthropic disputes the administration's targeting of its models, arguing that the capability identified is not unique and can be demonstrated in rival systems, including those developed by OpenAI.Financial Times
  • The government's freeze does not apply to leading models from OpenAI, despite those having demonstrated comparable capabilities.Financial Times
  • The Pentagon has named Anthropic as a supply-chain risk to national security, and the two sides are in litigation over the designation.Financial Times
Article not found
⌂ CuriousCats.ai

Article

‹

Source Citations