NewsBite analysis · AI, Regulation, Security
The Hindu
U.S. allows limited access to Anthropic’s Mythos AI model
Anthropic has received authorisation to allow a limited number of American cybersecurity firms to access its AI model Mythos 5, which had been restricted due to national security concerns. The U.S. government had previously mandated a halt to access, citing vulnerabilities in the model's safeguards. This action has been viewed by some as government overreach. Secretary of Commerce Howard Lutnick commended Anthropic's collaboration in addressing associated risks. There is ongoing tension as Anthropic has resisted government requests for its technology to be used in mass surveillance or military applications, straining its relations with government agencies.
MonitorConfidence: MediumShort term horizonCybersecurity risk: High
Published 27 Jun 2026, 07:35 UTC2 min read
Visit original article ↗Lead image — The Hindu
Summary
TL;DR
- Anthropic's Mythos AI model access was previously blocked by the U.S. government.
- Access is now allowed for a select group of cybersecurity firms.
- Concerns about national security prompted the initial block.
- Anthropic is seeking to expand access to its AI models.
- The relationship between Anthropic and the U.S. government remains contentious.
Interesting facts
- The U.S. government previously suspended access to Mythos AI due to security vulnerabilities.
- Anthropic's refusal to support surveillance and military applications led to contract cancellations with the Pentagon.