Open Source

All highly capable AI models should be regulated through pre-deployment and ongoing testing. But any safety regime that works for domestic models should work for Chinese ones.

Software ecosystems are strategic infrastructure, yet they remain almost invisible within national security risk assessments.

Digital open source data can be ethically deployed to strengthen investigations and prosecutions on gender crimes in Iran and elsewhere.

Retrieval-Augmented Generation offers federal agencies a way to build AI systems today within legal frameworks that already exist.

Technology can help bring justice for Ukrainian survivors of sexual and gender-based crimes, but the process is not without challenges.

China's accelerating wave of AI regulation makes one thing clear: the era of lightly-governed open-source AI in the country is ending.

The focus on open source investigation of serious international crimes often comes at the expense of more effective local expertise.

Policymakers should balance the benefits of open-source AI development with the need to address geopolitical considerations.

Brianna Rosen, Lennart Heim, Keegan McBride, & Lauren Wagner discuss what DeepSeek means for national security and U.S. policy going forward.

Beyond the hype, a crucial look at the real costs of China's DeepSeek, why compute still matters, and what US policymakers should do about it.

Trump's transactional approach to foreign policy could prove conducive to breakthrough AI agreements — especially with China.

Critics of open source AI must consider the security implications of strategic competition with China.