A nonprofit that educates policymakers, civil society, and the public about AI capabilities and dangers through interactive live software demonstrations.
A nonprofit that educates policymakers, civil society, and the public about AI capabilities and dangers through interactive live software demonstrations.
People– no linked people
Updated 05/18/26Funding Details
Updated 05/18/26- Annual Budget
- $976,565
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
Org Details
Updated 05/18/26The Civic AI Security Program (CivAI) is a 501(c)(3) nonprofit founded in 2023 by Lucas Hansen and Siddharth Hiregowdara, both software engineers who previously worked together at Qualia, a real estate software company Hansen co-founded. CivAI was launched in the wake of ChatGPT's late 2022 debut, motivated by the urgency of helping society understand and respond to rapidly advancing AI capabilities. Rather than producing academic research papers, CivAI builds interactive software experiences that give diverse audiences a visceral, hands-on sense of what AI can do and where it is headed. Their demonstrations cover AI-powered cybersecurity threats (phishing, voice cloning), deepfakes, election security risks, misinformation generation, elder fraud scenarios, and biological risks. Notable public demos include The Future of Phishing, Giving AI the Rorschach Test, the Deepfake Sandbox, and an AI bioweapons demonstration that drew significant attention in Washington. Since founding, CivAI has delivered over 100 briefings to a wide range of audiences. Government briefings have included NIST staff at the U.S. Department of Commerce, a multistate working group of state lawmakers studying AI regulation, the Texas House select committee on artificial intelligence, Arizona election officials across 15 counties, Washington D.C.-area law enforcement, Congressional offices and committees, and national security officials. Civil society audiences have included AARP and other organizations. CivAI also presented at the AI Action Summit in Paris in February 2025. In early 2024, CivAI became a founding member of the Artificial Intelligence Safety Institute Consortium (AISIC) established at NIST. Their bioweapons demonstration, which showed how popular AI models could be coaxed into providing step-by-step instructions for creating dangerous pathogens, was featured in a major TIME article in January 2026 and led to private briefings with lawmakers and national security officials. CivAI's work has been covered extensively in national media, including the New York Times, the Washington Post, TIME, CNN, ABC News, FOX, TheStreet, GovTech, and Pluribus News. The organization uses purely open-source and commercially available AI tools in its demonstrations, emphasizing that these capabilities are already accessible to bad actors. Lucas Hansen brings 10 years of technology experience and holds a B.A. in computer science from Stanford University. He previously served as CTO of Qualia. Siddharth Hiregowdara held various product management and engineering roles at Qualia and holds a B.A. from Dartmouth College.
Theory of Change
Updated 05/18/26CivAI believes that policymakers, government officials, and the public lack a visceral understanding of what AI systems can actually do, which leads to inadequate regulation and preparedness. By building interactive software demonstrations that concretely show AI capabilities and risks - from deepfakes to bioweapons instructions to sophisticated phishing - and presenting these directly to decision-makers, CivAI aims to create a foundational education that enables better-informed policy decisions. The causal chain runs from hands-on demos to improved understanding among key stakeholders to more effective governance and accountability for AI developers, ultimately reducing the risk of harm from misused AI capabilities.
Grants Received
Updated 05/18/26Projects– no linked projects
Updated 05/18/26Discussion
Key risk: By centering briefings on near-term misuse (phishing, deepfakes, step-by-step bio instructions), they may redirect scarce policy attention toward downstream content threats over upstream frontier governance, limiting counterfactual impact on existential risk.
Case for funding: CivAI has unusual access to U.S. policymakers and AISIC, and their high-fidelity interactive demos (including the widely covered bioweapons prompt tests) uniquely translate abstract frontier-model dangers into visceral understanding that catalyzes demand for evaluations, standards, and capability controls.