Seth Lazar is a philosopher working on AI ethics, the philosophy of AI and computing, and democratic theory. He is Professor at the Johns Hopkins University School of Government and Policy and principal investigator of the Machine Intelligence and Normative Theory (MINT) Lab, which he founded at the Australian National University (Australian National University (ANU)) and which has since relocated to Johns Hopkins (Source: mintresearch.org). The lab describes its work as using philosophical and computational research to chart a course through the AI transition and to help design institutions that preserve liberal democratic values.
Work on AI ethics
Lazar is the author of Anticipatory AI Ethics (Knight Columbia, May 1, 2026), the essay that names and defends the technological horizon as a constraint on anticipatory AI ethics (Anticipatory AI Ethics (Lazar, Knight Columbia, May 1 2026); see Anticipatory AI Ethics (Lazar framework, including the Technological Horizon)). The framework advances conditional hazard and opportunity identification as the methodological frame for anticipatory work. The essay defends reasoning about frontier systems before harms fully materialize while addressing objections that such reasoning is speculative or misdirected (Source: mintresearch.org). Lazar's earlier essay with Nelson (2023) applied anticipatory ethics to bioethics as a companion treatment; his co-author on that piece is Alondra Nelson, with whom he published in Science.
His prior research cited in Anticipatory AI Ethics addresses platform power and agentic AI. This includes Lazar et al. (2024) on platform power and agentic AI; Chan et al. (2025), co-authored work on agentic AI; and Kapoor et al. (2025) on the platform economy and agentic AI, co-authored with Sayash Kapoor and Arvind Narayanan. Lazar's work engages with the AI as Normal Technology framing (Narayanan-Kapoor 2025) and with AI Existential Risk discourse.
Political philosophy of algorithmic power
Lazar's central line of work treats automated systems as exercising a kind of governance, and asks what democratic legitimacy, authority, and accountability require when administration becomes computational. He delivered the Stanford Tanner Lectures on AI and Human Values in January 2023 in two parts, "Governing the Algorithmic City" and a second lecture on communicative justice and the political philosophy of language models (Source: philosophy.stanford.edu). The first was developed into "Governing the Algorithmic City," published in Philosophy & Public Affairs in 2025 (Source: onlinelibrary.wiley.com), which argues that algorithmic governance raises new challenges for political philosophy concerning the justification of authority and the foundations of legitimacy (Source: uchv.princeton.edu). Related work argues that AI can become an authority-like structure shaping options, attention, and social relations, and delivered the Kamm Lecture on the nature and justification of algorithmic power in 2022 (Source: mintresearch.org).
A separate strand critiques the idea of a general right to an explanation for opaque algorithmic decisions, separating accountability demands from a blanket explanatory entitlement, and analyzes targeted AI-driven influence by locating objections in autonomy, fairness, and communicative justice rather than treating persuasion as automatically wrongful (Source: mintresearch.org). A more recent preprint argues that normative competence is central to AI alignment and governance, reviews evidence that frontier models already display substantial moral reasoning abilities, and outlines methods for improving and evaluating them (Source: mintresearch.org); see Normative Competence.
Editorial and advisory roles
Lazar co-edited The Political Philosophy of Data and AI (May 2022) with Kate Vredenburgh and Annette Zimmermann, and a Philosophical Studies special issue on normative theory and AI with Claire Benn, Todd Karhu, and Pamela Robinson. He joined the Knight First Amendment Institute as senior AI adviser in January 2024 (Source: mintresearch.org). MINT Lab has convened workshops on democracy and AI, including work on how democratic institutions can be redesigned to remain effective against advanced AI systems.
Relationships
- affiliated-with: Australian National University (ANU) (MINT Lab founding home), Knight First Amendment Institute (Knight Columbia) (senior AI adviser since January 2024; Knight Columbia 2026 symposium author)
- co-author-with: Sayash Kapoor, Arvind Narayanan (Kapoor et al. 2025), Alondra Nelson (Science)
- engaged-with: AI as Normal Technology (Narayanan-Kapoor 2025), AI Existential Risk discourse
- author-of: Anticipatory AI Ethics (Lazar, Knight Columbia, May 1 2026)
- supports: Anticipatory AI Ethics (Lazar framework, including the Technological Horizon), Normative Competence
Sources
Confidence
Medium. The affiliation change from ANU to Johns Hopkins and the publication list come from the MINT Lab site, a first-party source, corroborated on the Stanford, Princeton, and Wiley pages for the Tanner Lectures and the Philosophy & Public Affairs article. The date of the Johns Hopkins move is not stated on the source consulted.