Partnership for Public Service Event

At the AI Center for Government’s “How Government is Using AI Now” panel on 26 February 2026, our founder Adrianna Tan shared practical lessons from working on the Department of Defense's groundbreaking medical LLM red teaming project, in collaboration with Humane Intelligence.

She was joined by Alexander Schneider (Colorado Digital Service) and Martin Skorczynski (Government Accountability Office) in discussing LLMs as one tool among many within the government digital toolkit: potentially valuable when aligned with mission needs but needing rigorous evaluation before deployment.

Structured adversarial testing, or red teaming, uncovers failure modes that might have otherwise been invisible, or worse, materialize only when harm reaches a real person or group. More than 200 people attended this virtual discussion facilitated by the Partnership for Public Service.

The Q&A also covered how key government technology leaders in state and federal government are thinking about AI use, including use cases and potential harms and risks.

Adrianna shared an interest in understanding how government technology roles will evolve to fit the new landscape at both the executive and non-executive level, acknowledging that there are now Chief AI Officers alongside CIOs and CISOs. How can government technology hiring and talent acquisition evolve to meet new needs? She also raised the question of who is accountable in government if government deploys AI that causes harm to marginalized groups, referencing recent California legislation on this matter.