ACLU to FOIA information about national security uses of AI
The ACLU will be seeking information about how the government is using artificial intelligence in national security, Axios has learned.
Why it matters: The development of AI has major implications for security, surveillance, and justice. The ACLU's request may help shed some light on the government's often opaque applications of AI.
Driving the news: Later today the ACLU will be filing a broad Freedom of Information Act (FOIA) request to the CIA, the NSA, the Department of Homeland Security and other agencies concerning the government's use of AI, especially in the area of national security.
- "The problem with these AI systems is that they're black boxes," says Patrick Toomey, senior staff attorney at the ACLU National Security Project. "The public needs to know exactly what kinds of fundamental decisions about our lives the government is handing over to AI."
Details: The ACLU is specifically concerned about "vetting and screening processes in agencies like Homeland Security, and tools that can analyze voice, data and video," says Toomey.
- Another area of concern is the possibility that AI systems could be "biased against people of color, women and marginalized communities," he adds.
- "AI systems could be used to supercharge government activities to unfairly scrutinize communities through intrusive surveillance, questioning and even detention and watchlisting."
Background: The FOIA request was prompted in part by a recent 750-page report put out by the National Security Commission on Artificial Intelligence that lays out a case for the U.S. to embrace AI throughout the national security sector.
- Much of the report views AI through the prism of competition with China, which has used artificial intelligence to help create a vast surveillance state.
- "We don't have to go to war with China," former Google CEO and commission chair Eric Schmidt told my Axios colleague Ina Fried when the report was released. "We do need to be competitive."
The bottom line: "AI will be a society shaping technology," says Toomey.
- "Because of that, we have to be considering what safeguards and protections are necessary from the start."