Center for Democracy & Technology
AI-generated text. This page was generated using artificial intelligence.
Center for Democracy & Technology (CDT) is a nonprofit organization working on civil rights and civil liberties in technology policy. Its AI work includes accountability, discrimination, privacy and the practical design of governance processes. Its AI Governance Lab was established in October 2023 to contribute public-interest expertise to debates over auditing, safety and mitigation of harms.[1]
Role in AI law and policy
CDT connects AI governance with established questions about how institutions make consequential decisions and how people can challenge them. Its AI program encompasses employment and insurance decisions, government services, law-enforcement uses, online speech and information integrity. This gives its work a scope extending beyond the safety of the most capable general-purpose models to systems already used in daily administration and commerce.[2]
The program also maintains a searchable collection of its AI-policy work. Its description emphasizes access to positions across time, including positions that may have changed. For a reader following a particular bill or regulatory proposal, the relevant document and its date therefore matter more than a broad characterization of CDT as supporting or opposing AI regulation in general.[2]
AI Governance Lab
The launch announcement named Miranda Bogen as founding director and Kevin Bankston as senior advisor. It described engagement with companies, multistakeholder initiatives, other advocates and policymakers. Initial philanthropic support included the Ford Foundation, Open Society Foundations and Siegel Family Endowment. These are details of the 2023 launch, rather than a certification that the same staffing or funding arrangements continue unchanged.[1]
Independent reporting by IAPP described the lab's objective as translating high-level commitments into workable standards and practices. The article reported Bogen's concern that implementation details determine whether accountability measures actually work. It also situated the founders' experience in industry within their move to public-interest advocacy. This supplies outside context for the lab's approach without establishing that its proposals have been adopted by a regulator.[3]
Assessments, audits and independence
CDT's January 2025 report Assessing AI distinguishes activities often grouped together as auditing or evaluation. It identifies different purposes: learning about a system, judging the adequacy of safeguards, communicating impacts and changing behavior. The report argues that an assessment should be designed around its intended purpose rather than selected because an accountability label sounds reassuring.[4]
The analysis compares both scope and independence. A broad exploration of possible harms differs from a narrow benchmark test; an external assessment can still be constrained if the developer chooses the questions or controls disclosure. CDT recommends transparency about methods, assumptions and the system version examined, together with a range of participants and approaches. The report does not suggest that any single audit can demonstrate safety or legality for every use of a system.[4]
Funding and governance disclosures
CDT publishes donor information, tax filings and financial reports. Its page updated July 1, 2026 identifies 2025 revenue from foundations, corporations, events, individuals and other sources. It expressly labels the displayed 2025 financial statements unaudited. The disclosure lists technology companies among supporters, and some contribution categories include pro bono services rather than only cash.[5]
CDT states that supporters have no control over its projects, priorities or work product and links its gift-acceptance policy. This is an institutional independence rule. A donor listing establishes disclosed support, not that a donor wrote a policy recommendation, approved a publication or financed a particular AI project. The launch funding of the AI Governance Lab should likewise be distinguished from CDT-wide revenue.[5]
Selected contributions
- AI Governance Lab launch (October 2023), bringing technical and policy expertise into civil-society participation.[1]
- Assessing AI (January 2025), a framework for matching assessment methods to accountability goals.[4]
- AI-policy research collection, covering both current positions and their development over time.[2]
Related articles
References
- ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 CDT, CDT Launches New AI Governance Lab Headed by Leading Experts, October 23, 2023. Accessed September 12, 2026.
- ↑ 2.0 2.1 2.2 CDT, AI & Machine Learning. Accessed September 12, 2026.
- ↑ Caitlin Andrews, Leaders of CDT’s new AI Governance Lab want to shape policy dialogue, IAPP, November 1, 2023. Accessed September 12, 2026.
- ↑ 4.0 4.1 4.2 Miranda Bogen, Assessing AI: Surveying the Spectrum of Approaches to Understanding and Auditing AI Systems, January 16, 2025. Accessed September 12, 2026.
- ↑ 5.0 5.1 CDT, Financials, updated July 1, 2026. Accessed September 12, 2026.