The U.S. government has lifted restrictions on overseas access to Anthropic's high-performance artificial intelligence (AI) models, Phaeble 5 and Mythos 5.

Reuters Yonhap News

Reuters Yonhap News

View original image

On June 30 (local time), Anthropic announced via X (formerly Twitter), "We have received notification from the U.S. Department of Commerce that export controls on Claude Phaeble 5 and Mythos 5 have been lifted," adding, "We will begin restoring access starting tomorrow and will share further updates soon."


The export controls were lifted after approximately three weeks. On June 12, the U.S. Department of Commerce imposed export control measures on Anthropic through a confidential letter. As a result, Anthropic was required to obtain prior approval from the U.S. government, regardless of the foreign national's location, if it wished to allow foreign nationals access to Claude Phaeble 5 or Mythos 5, or to provide these models to any region worldwide. In response, Anthropic deactivated its high-performance AI models Mythos 5 and Phaeble 5.


The key to lifting the export controls was convincing White House officials that Anthropic had established sufficient safeguards to prevent malicious actors from circumventing the models' safety features. In a letter confirmed by Bloomberg, U.S. Secretary of Commerce Howard Lutnick stated that Anthropic had pledged to "proactively deter and respond to" security risks associated with the models.


The U.S. government is tightening oversight of the release of new AI models. This comes amid growing concerns that advanced AI models, which are fueling the AI boom and driving large-scale investments, could be misused by military and intelligence authorities in countries of concern such as China and Russia. Previously, Bloomberg reported that the U.S. government had also requested OpenAI to delay the full public release of GPT-5.6 and to make it available first only to a select group of trusted partners.



Meanwhile, Anthropic has launched its scientific research AI platform, Claude Science. Claude Science is designed to help scientists streamline research processes, analyze data, and manage complex computational workflows. Anthropic explained that its goal is to enable researchers to conduct their work within a single environment without having to move between multiple databases and tools.


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing