US relaxes export controls on Anthropic AI, opening limited access to its Mythos 5 model
Executive summary: The Trump administration partially lifted export controls on Anthropic AI, permitting selected US entities to access the Mythos 5 model while keeping a more advanced model blocked. This policy shift affects the availability of cutting‑edge AI for US commercial and defense sectors, influences the global AI chip supply chain, and signals a potential softening of US export control stance amid US‑China tech rivalry.
Who is involved: Trump administration (particularly the Department of Commerce and Bureau of Industry and Security), Anthropic AI, US companies and government agencies eligible for licences, allied governments monitoring the change.
Likely next: Further reviews of the remaining restrictions, possible issuance of additional licences for the blocked model, and ongoing dialogue with allies on coordinated AI export control frameworks.
The Trump administration has decided to partially lift the export ban imposed on Anthropic’s most advanced AI models, allowing a select group of US companies and government agencies to procure and use the Mythos 5 system. A second, more capable Anthropic model remains under restriction, indicating a calibrated approach rather than a full reversal. The move reflects shifting US policy on AI technology transfer amid growing strategic competition with China and pressure from domestic tech firms.
Timeline
- — Trump administration partially lifts Anthropic’s AI export ban (Politico Europe)
Analysis — what this means
Likely next events
- Potential full lift of the remaining Anthropic AI export ban
- Responses from allied governments reassessing their own AI export controls
- Increased procurement of Anthropic’s Mythos 5 model by US defense contractors and cloud providers
Sectors affected
- Artificial Intelligence
- Semiconductors
- Defense and Aerospace
Regulatory implications
- Re‑evaluation of Entity List restrictions for AI firms
- Possible revisions to the Export Administration Regulations (EAR) concerning AI software
- Enhanced multilateral coordination on AI technology transfer rules
Historical parallels
- 2020 selective easing of Huawei sanctions via temporary licences
- 2021 removal of certain AI software bans for academic collaboration under the US‑EU Trade and Technology Council
- Prior relaxations of export controls on high‑performance computing chips for firms like Nvidia and AMD
Key entities
Sources
- Trump administration partially lifts Anthropic’s AI export ban — Politico Europe
Related cases
- Anthropic AI test showed autonomous phishing capability, raising cyber‑risk concerns for AI deployment
- The Trump administration’s partial lift of Anthropic’s AI export ban opens limited access to its Mythos 5 model while keeping a more advanced model under restriction
- Chip‑stock rally rebounds on Iran diplomacy and Anthropic AI clash
- U.S. export curbs on Anthropic AI software heighten German digital sector's reliance on American technology
- U.S. curbs on Anthropic AI models boost Zhipu's valuation as investors view it as a beneficiary of tighter AI regulations
- US block on Anthropic AI software heightens German digital sector's reliance on US technology