Center for Humane Technology

The Center for Humane Technology (CHT) is a San Francisco-based nonprofit founded in 2018 by former Google Design Ethicist Tristan Harris, Aza Raskin, and Randima Fernando — three technologists who concluded that the attention-extraction incentives of the major digital platforms were structurally incompatible with democratic society and individual well-being. CHT's core strategy is counter-narrative public engagement: reframing how non-specialist publics, policymakers, and media understand technology's design incentives, through documentary participation, podcast, live briefing, voter-education work, and youth curriculum. With approximately 30 staff as of 2026, it operates as a 501(c)(3) nonprofit funded by the Ford Foundation, Omidyar Network, Knight Foundation, Open Society Foundations, and Craig Newmark Philanthropies, among others.

Founding and the Time Well Spent movement

Harris, while at Google, created a 2013 internal presentation — "A Call to Minimize Distraction & Respect Users' Attention" — that circulated widely inside Google and articulated the core CHT argument: that platforms designed to maximize engagement were actively working against users who intended only to use them briefly. This framing became the basis for the "Time Well Spent" movement, a public campaign calling on technology companies to redesign products around user well-being rather than attention capture. Aza Raskin — who had invented the infinite-scroll interface as a developer and later came to view it as a compulsive-design artifact — joined as co-founder, and Randima Fernando joined as co-founder and COO. CHT received its IRS tax-exempt ruling in 2018. Harris, Raskin, and Fernando collectively represented a rare profile in AI-harm advocacy: technologists who had built the systems they were now criticizing from the inside, and whose insider knowledge gave their framing unusual credibility with both press and policymakers.

The Social Dilemma and AI Dilemma — narrative strategy at scale

CHT's primary public-engagement vehicle through 2020 was the Netflix documentary The Social Dilemma (directed by Jeff Orlowski-Yang, Exposure Labs), in which Harris, Raskin, Fernando, and a roster of former social-media engineers testified to platform design's deliberate manipulation of attention and the resulting harms to mental health, political discourse, and democratic capacity. CHT worked closely with Orlowski-Yang on the film and it lists on CHT's own impact timeline as the organization's central public-engagement milestone. The Social Dilemma reached more than 100 million viewers in 190 countries across 30 languages and won two Emmy Awards. In 2023, a coalition of 42 state Attorneys General cited The Social Dilemma as a catalyst for their lawsuit against Meta for designing addictive features that harm children.

In March 2023, as generative AI entered mass deployment, Harris and Raskin delivered The AI Dilemma — a private presentation to leading technologists and decision-makers that applied CHT's social-media analysis framework to the new AI risk landscape. The presentation was subsequently released publicly, went viral, and became the basis for CHT's congressional briefings. Harris and Raskin testified before Senate and House committees on algorithmic harms and AI risk, deploying the same presentation logic — AI companies are caught in a race to deploy without adequate safety measures; the social media precedent shows what happens when this race wins. The AI Dilemma was CHT's formal entry into AI-specific advocacy from its earlier social-media focus.

Your Undivided Attention — ongoing public reframing

Your Undivided Attention is CHT's primary ongoing public-engagement channel — a podcast co-hosted by Harris and Raskin that by 2024 had accumulated more than 25 million downloads. The podcast carries CHT's counter-narrative framing of technology's incentive structures to a general-public audience, bringing in technologists, policymakers, and researchers to contextualize AI and digital-platform design for listeners who are not technical specialists. Topics span AI risk, social-media design harms, democracy and disinformation, children and technology, and the structural incentives driving AI development. At the AI policy layer, the podcast has served as a vehicle for briefing the non-specialist movement adjacent public — parents, educators, concerned citizens — on the regulatory and technical dynamics they are not otherwise equipped to parse.

Teaching for Humane Technology — youth education program

CHT's Teaching for Humane Technology program provides a free youth toolkit for students and educators at ages 13-25, combining issue guides (how persuasive technology works, the financial incentives behind platform design) with action guides (goal-setting for technology use, storytelling from personal social media experience, and imagining possibilities for humane technology). The toolkit is designed for high-school classrooms, disseminated freely to educators and students, and intended to build AI- and platform-literacy directly into secondary school curricula. It represents CHT's most direct educational engagement with young non-specialists — the population most heavily targeted by the attention-extraction systems CHT's public work critiques.

Humane Evals — measuring AI's psychosocial impact

In 2025-2026, CHT launched Humane Evals, a research initiative focused on measuring AI's psychological and social impacts on users. Led by researchers Imran Khan and Jared Moore, the program addresses the gap between the proliferation of technical AI benchmarks (testing AI competence on specific tasks) and the near-absence of evaluations measuring AI's effects on human well-being, reasoning, and psychology. The initiative is developing a prototype evaluation framework for what it terms AI's "psychosocial" impacts — attentional, emotional, and relational effects on users — that technical benchmark regimes do not capture.

2026 electoral voter guide

For the 2026 US midterm elections, CHT published an AI and the 2026 Midterms voter guide designed to help non-specialist voters evaluate candidates on AI-related positions: whether candidates address AI's impact on jobs and local communities, support guardrails on AI development, ensure community voice in AI data-center siting decisions, and have concrete plans for workers affected by AI-driven job displacement. The guide evaluates candidate framing — flagging candidates who frame AI only in terms of "national competitiveness" without addressing local consequences as weaker on the guide's criteria. It represents CHT's explicit entry into electoral advocacy, applying the voter-guide model standard to AI governance as a voting issue.