Jennifer King is the Privacy and Data Policy Fellow at the Stanford Institute for Human-Centered Artificial Intelligence (Stanford HAI) (Regulating Data Brokers in the Age of AI: A California Case Study). Her work addresses AI privacy questions, including training-data privacy, consumer-AI privacy frameworks, the data-broker market, and the relationship between federal and state AI privacy regulation.
Background
King holds a PhD in Information Management & Systems from UC Berkeley. She works at Stanford HAI (Stanford HAI), where her research focuses on privacy and data policy as they apply to AI systems, and she is also affiliated with Stanford Law School's Center for Internet and Society (Source: papers.ssrn.com; hai.stanford.edu).
Research and publications
- Government data and algorithmic fairness. King co-authored, with Daniel E. Ho, Russell C. Wald and Christopher Wan, Building a National AI Research Resource: A Blueprint for the National Research Cloud (Stanford HAI, October 2021), and, with Ho, Arushi Gupta, Victor Wu and Helen Webley-Brown, "The Privacy-Bias Tradeoff: Data Minimization and Racial Disparity Assessments in U.S. Government" (ACM FAccT 2023). Both are cited in essays collected in The Digitalist Papers (Stanford, Volumes 1–2).
- Rethinking Privacy in the AI Era (2024). With Caroline Meinhardt, King wrote Rethinking Privacy in the AI Era: Policy Provocations for a Data-Centric World, a Stanford HAI white paper dated February 22, 2024. Nita Farahany's "three privacy problems" framework adapts it, alongside Ryan Calo's July 2024 Senate testimony; see Three Privacy Problems AI Creates (Source: hai.stanford.edu). The white paper is also cited in the governance-options study Regulating Under Uncertainty: Governance Options for Generative AI (G'sell, 2024).
- Data privacy and foundation models (2026). With Tiffany Saade, King wrote "Data Privacy and Foundation Models: Can We Have Both?", a Stanford HAI Policy & Society Issue Brief released in April 2026. It argues that privacy risk in foundation models is structural, arising separately at each stage of the model lifecycle from pre-training scraping of personally identifiable information to memorization and regurgitation in outputs and the "intimate data" users disclose to chatbots, and that frameworks including the GDPR, CCPA and HIPAA do not fully capture these risks (Data Privacy and Foundation Models: Can We Have Both? — King & Saade (Stanford HAI Issue Brief, April 2026)).
- Data brokers (2026). With Anna-Maria Gueorguieva, Apoorva Panidapu and Daniel E. Ho, King co-authored Regulating Data Brokers in the Age of AI: A California Case Study, a Stanford HAI and RegLab brief condensing their FAccT 2026 paper "Privacy Without Remedy." Reviewing all 522 brokers registered with the California Privacy Protection Agency in 2025, it found that 45% reported no information about consumer requests received and 9.2% were fully compliant, reporting on all four request types, and it recorded that 32 brokers in the 2026 registry sell data to generative-AI developers (Regulating Data Brokers in the Age of AI: A California Case Study).
Positions and statements
On government surveillance, King contributed "Why Anthropic's Rejection of Pentagon Deal Reveals AI's Privacy Crisis" to a Stanford HAI roundtable on the Anthropic–Pentagon dispute (Who Decides How America Uses AI in War? — Stanford HAI faculty roundtable (2026)). She argues that large language models, trained on personal information from public and semi-public websites, public records and chatbot conversations, lower the bar for individual surveillance, and that "what stand between us and their usage for profiling and surveillance are the companies' choices to not cross this line." She describes Anthropic's refusal to allow its tools to be used for mass domestic surveillance as the first explicit public confirmation that such tools can be put to those uses, notes that Palantir has contracted with the US government to provide services that large commercial providers are reluctant to allow, and leaves open whether society is content to leave that discretion to the companies (Source: hai.stanford.edu).
On training-data privacy, King has written on the application of US state privacy law to AI training-data sourcing, work relevant to the Bartz v. Anthropic and Hachette et al. v. Meta (and Mark Zuckerberg) cases.
On consumer-AI privacy, she has addressed consumer-AI privacy disclosure questions, which bear on the inference-time-privacy dimension of AI and Privacy.
On the federal-versus-state question, her Stanford HAI publications address federal AI privacy preemption, relevant to the context of Executive Order 14365.
Relationships
- related: AI and Privacy, AI Content Licensing.
- related: Data Privacy and Foundation Models: Can We Have Both? — King & Saade (Stanford HAI Issue Brief, April 2026), Toward Responsible AI in Health Insurance Decision-Making — Mello, Trotsyuk, Djiberou Mahamadou, Char (Stanford HAI Policy Brief, February 2026).
- related: Stanford HAI.
- related: Regulating Data Brokers in the Age of AI: A California Case Study (co-author), Who Decides How America Uses AI in War? — Stanford HAI faculty roundtable (2026) (contributor), Three Privacy Problems AI Creates (adapts her 2024 white paper), Tiffany Saade (co-author).
- related: AI and Surveillance — her argument that developers' usage policies are the operative limit on LLM-enabled profiling.
Sources
- Data Privacy and Foundation Models: Can We Have Both? — King & Saade (Stanford HAI Issue Brief, April 2026)
- Regulating Data Brokers in the Age of AI: A California Case Study
- Who Decides How America Uses AI in War? — Stanford HAI faculty roundtable (2026)
- The Digitalist Papers (Stanford, Volumes 1–2)
- Regulating Under Uncertainty: Governance Options for Generative AI (G'sell, 2024)
- Stanford HAI, Rethinking Privacy in the AI Era (2024): hai.stanford.edu