White House Pushes Back Against Anthropic’s Mythos Expansion
None
<p style="font-weight: 400;">The White House is opposing Anthropic’s plan to expand access to its Mythos AI model, creating a high-stakes confrontation between the U.S. government and a top AI developer about how leading-edge AI models can be distributed.</p><p style="font-weight: 400;">When Anthropic <a href="https://securityboulevard.com/2026/04/anthropic-unveils-restricted-ai-cyber-model-in-unprecedented-industry-alliance/">unveiled Mythos</a> on April 7, it allowed access to only a small cohort of companies, including Amazon, Microsoft, Google, and NVIDIA. The reasoning was that the model’s ability to identify and exploit software vulnerabilities is so great that it would imperil critical infrastructure if it was ever publicly released.</p><p style="font-weight: 400;">Anthropic has since proposed adding roughly 70 new organizations to the program, more than doubling the current user base.</p><h3 style="font-weight: 400;"><strong>Two Major Concerns</strong></h3><p style="font-weight: 400;">Administration officials have told Anthropic they oppose greater Mythos access based on two concerns: the model’s potential for misuse and the infrastructure required to support a wider rollout.</p><p style="font-weight: 400;">On the potential for misuse, Mythos has prompted heightened scrutiny inside government agencies responsible for critical infrastructure protection, and security incidents have reinforced the need for scrutiny. Shortly after Mythos was introduced in a limited release, unauthorized users were able to gain access through private channels. The breach exposed the difficulty of containing high-demand AI systems, particularly those with clear offensive applications.</p><p style="font-weight: 400;">At the end of March, Anthropic accidentally exposed the internal source code for its Claude Code agentic AI tool. The leak, caused by human error, raised troubling concerns about the company’s ability to contain sensitive information.</p><p style="font-weight: 400;">Anthropic has acknowledged the risks inherent in its technology, which have shaped the company’s decision to avoid a full public release, even as enterprise demand intensifies.</p><p style="font-weight: 400;">Officials are also evaluating whether Anthropic has sufficient computing capacity to support a larger pool of users. Expanding access without adequate compute power could degrade performance for government users, who are already accessing the system for security-related tasks. While Anthropic has secured partnerships to boost compute supply, those resources are still being assembled.</p><h3 style="font-weight: 400;"><strong>New Issues in Accessing AI Models</strong></h3><p style="font-weight: 400;">The relationship between Anthropic and the federal government has been uneven, at best. Earlier disputes tied to defense applications of Anthropic’s AI technology led to a breakdown in cooperation, after the company refused to remove guardrails that blocked its model from use for autonomous lethal weapons or mass U.S. surveillance. In response, Defense Secretary Pete Hegseth labeled Anthropic a “supply chain risk” to national security. That history likely continues to influence current negotiations, even as both sides maintain active discussions around the Mythos rollout.</p><p style="font-weight: 400;">Government officials are exploring more ways to integrate the company’s models into government workflows despite earlier restrictions. This approach of limiting external access while increasing internal use illustrates the crucial importance of maintaining access to advanced AI capabilities.</p><p style="font-weight: 400;">Indeed, as frontier AI systems grow more powerful, access is necessarily becoming conditional, shaped by new relationships that have not yet developed trust, infrastructure readiness, and alignment with national security priorities.</p><p style="font-weight: 400;">For enterprise buyers, this dispute introduces a new variable: regulatory acceptance may determine availability as much as technical capability. Adding complexity to an already complex challenge, procuring AI systems is moving from a technical decision to a policy-controlled process.</p><div class="spu-placeholder" style="display:none"></div><div class="addtoany_share_save_container addtoany_content addtoany_content_bottom"><div class="a2a_kit a2a_kit_size_20 addtoany_list" data-a2a-url="https://securityboulevard.com/2026/04/white-house-pushes-back-against-anthropics-mythos-expansion/" data-a2a-title="White House Pushes Back Against Anthropic’s Mythos Expansion"><a class="a2a_button_twitter" href="https://www.addtoany.com/add_to/twitter?linkurl=https%3A%2F%2Fsecurityboulevard.com%2F2026%2F04%2Fwhite-house-pushes-back-against-anthropics-mythos-expansion%2F&linkname=White%20House%20Pushes%20Back%20Against%20Anthropic%E2%80%99s%20Mythos%20Expansion" title="Twitter" rel="nofollow noopener" target="_blank"></a><a class="a2a_button_linkedin" href="https://www.addtoany.com/add_to/linkedin?linkurl=https%3A%2F%2Fsecurityboulevard.com%2F2026%2F04%2Fwhite-house-pushes-back-against-anthropics-mythos-expansion%2F&linkname=White%20House%20Pushes%20Back%20Against%20Anthropic%E2%80%99s%20Mythos%20Expansion" title="LinkedIn" rel="nofollow noopener" target="_blank"></a><a class="a2a_button_facebook" href="https://www.addtoany.com/add_to/facebook?linkurl=https%3A%2F%2Fsecurityboulevard.com%2F2026%2F04%2Fwhite-house-pushes-back-against-anthropics-mythos-expansion%2F&linkname=White%20House%20Pushes%20Back%20Against%20Anthropic%E2%80%99s%20Mythos%20Expansion" title="Facebook" rel="nofollow noopener" target="_blank"></a><a class="a2a_button_reddit" href="https://www.addtoany.com/add_to/reddit?linkurl=https%3A%2F%2Fsecurityboulevard.com%2F2026%2F04%2Fwhite-house-pushes-back-against-anthropics-mythos-expansion%2F&linkname=White%20House%20Pushes%20Back%20Against%20Anthropic%E2%80%99s%20Mythos%20Expansion" title="Reddit" rel="nofollow noopener" target="_blank"></a><a class="a2a_button_email" href="https://www.addtoany.com/add_to/email?linkurl=https%3A%2F%2Fsecurityboulevard.com%2F2026%2F04%2Fwhite-house-pushes-back-against-anthropics-mythos-expansion%2F&linkname=White%20House%20Pushes%20Back%20Against%20Anthropic%E2%80%99s%20Mythos%20Expansion" title="Email" rel="nofollow noopener" target="_blank"></a><a class="a2a_dd addtoany_share_save addtoany_share" href="https://www.addtoany.com/share"></a></div></div>