Software developers in the U.S. are utilizing AI models from China for coding tasks like writing, debugging, and reviewing. They are attracted by lower prices compared to American alternatives. However, these models pose significant security risks for U.S. software, according to a report by Booz Allen Hamilton. Booz Allen Hamilton's report specifically tested how these AI models respond when users appear to be affiliated with the U.S. government.
Software developers in the U.S. are utilizing AI models from China for coding tasks like writing, debugging, and reviewing. They are attracted by lower prices compared to American alternatives. However, these models pose significant security risks for U.S. software, according to a report by Booz Allen Hamilton.
Booz Allen Hamilton's report specifically tested how these AI models respond when users appear to be affiliated with the U.S. government. In May 2026, the firm conducted over 2,800 trials to assess these security implications.