The United States government has lifted the export control restrictions that had kept Anthropic’s most advanced artificial intelligence models blocked since mid-June. The Claude Fable and Claude Mythos models, considered the most powerful from the company that develops Claude (one of the most popular AI assistants in the world), are operational again after the Trump administration reconsidered its stance following weeks of intense negotiations.
The original decision to block these models stemmed from national security concerns: the government feared that such advanced capabilities could fall into the wrong hands or be used to develop weapons or cyberattacks. However, Anthropic argued that the measure was disproportionate and harmed the United States’ technological competitiveness against other countries such as China, which do not have this type of restriction. The final agreement includes voluntary security measures, such as the obligation to inform the authorities if any dangerous use of the models is detected.
For the average user, this decision means that AI assistants such as Claude will continue to improve and offer new capabilities without the limitations the block would have imposed. It also sets an important precedent for how governments balance national security with technological progress, a debate that will become increasingly frequent as artificial intelligence grows more powerful. If Anthropic had lost access to its most advanced models, the development of new features for users would have slowed considerably.
— Marta, for intermittent intelligence.
