The recent tensions between the U.S. Department of Defense and AI firm Anthropic have spotlighted a critical concern: Can the government legally conduct mass surveillance of American citizens using artificial intelligence? This issue, which gained traction following Edward Snowden’s revelations about NSA practices, remains complex and unresolved. The Pentagon’s interest in utilizing Anthropic’s AI technology, specifically Claude, to analyze extensive commercial datasets has raised eyebrows, particularly after Anthropic insisted that its AI not be involved in domestic surveillance or autonomous weaponry.
In a surprising turn, after negotiations with Anthropic failed, the Pentagon labeled the company as a supply chain risk— a designation typically reserved for foreign entities deemed a national security threat. Meanwhile, rival AI company OpenAI, which developed ChatGPT, secured a contract with the Pentagon allowing its AI to be used for all lawful purposes. Critics argue this could potentially open avenues for domestic surveillance, leading to public backlash against OpenAI, prompting the company to revise its agreement to explicitly prevent its technology from being used for such purposes.
The legal landscape surrounding AI surveillance is murky. While the Fourth Amendment offers protections against unreasonable searches and seizures, many forms of data collection that the government engages in do not qualify as surveillance under existing legal definitions. For instance, publicly available information—such as social media activity and voter registration details—can be freely accessed. Moreover, the government can purchase commercial data containing sensitive personal information, raising concerns about privacy. As technology evolves and AI capabilities expand, the law has struggled to keep pace, resulting in a scenario where the government can leverage massive amounts of data for surveillance without clear restrictions. This situation prompts ongoing debates about national security interests versus individual privacy rights, as officials argue that targeted intelligence gathering is necessary for counterintelligence and national safety.
Source: Is the Pentagon allowed to surveil Americans with AI? via MIT Technology Review