The Department of Homeland Security announced Monday that it's expanding its use of generative artificial intelligence with a trio of new pilot projects.

Why it matters: Like many government agencies, DHS is trying to figure out where it can harness generative AI despite risks related to accuracy, bias and other issues.

Driving the news: The agency announced three new efforts across its work.

• Investigators will be able to use open source large language models to more quickly summarize investigative reports and to improve the process of searching through reports. The agency hopes the pilot program will improve detection of fentanyl-related networks and help identify perpetrators and victims of child exploitation.
• Another project aims to use generative AI to improve training of immigration officers through more personalized materials while covering the latest legal issues and policies.
• Under FEMA, DHS hopes to help local communities by easing the time-consuming hazard mitigation planning process. The pilot could allow more communities to submit grants and gain funding to improve resilience and reduce disaster risks.

The big picture: DHS, like other agencies, is under pressure to increase its use of AI, while civil liberties groups are pushing to ensure privacy and equity.

• The agency held a recruiting event in Silicon Valley last month to add at least 50 artificial intelligence experts to its new "AI Corps" modeled on the U.S. Digital Service.

What to watch: I'll be interviewing Homeland Security Secretary Alejandro Mayorkas at Axios' What's Next Summit on Tuesday, where I plan to explore these new projects and the broader issues around the agency's use of AI.