Center for Humane Technology cuts staff to put founders in front

A narrower mission

CHT’s board says the founders’ ability to reach broad audiences and influential decision-makers is where the organization can make the greatest difference. It has not said how many employees will lose their jobs.

Board chair Premal Shah told WIRED the center needed to focus on one strategy: the one it had always been strongest at, as AI risks accelerated.

The shift leaves behind work that gave CHT a role in public policy and legal disputes. The organization helped organize congressional hearings on chatbot risks, advised participants in lawsuits and investigations into AI labs, and supported a major petition calling for a pause in AI development. It will no longer help with litigation or evaluate AI tools.

That change matters in part because CHT’s work has included scrutiny of companies developing the technology. Among the cases it advised on is an ongoing lawsuit against OpenAI, filed by the parents of a teenager who died by suicide after interacting with ChatGPT.

Founded in 2017, CHT has been competing for funding and policy influence with established nonprofits and a growing number of newer groups. Cyberattacks carried out by AI agents from OpenAI and other labs have added to calls for technology developers to be held responsible for design and development failures.

The center’s annual budget is about $6 million, with roughly the same amount in reserve. It is funded by wealthy families and foundations. Former employees worry that giving Harris a more prominent role will pull resources away from issues they consider more urgent, including AI’s effects on democracy and civil liberties.

The reach — and its limits

Harris has spent years making the case that technology companies should take responsibility for the social effects of their products. CHT’s early campaigns addressed political radicalization and harm to teenagers’ mental health. His warnings spread through Silicon Valley, and his national profile grew after Netflix released The Social Dilemma in 2020.

Harris founded CHT with Aza Raskin, who describes himself as the inventor of infinite scroll, and Randy Fernando, formerly an educator for developers at Nvidia. Harris has also warned that software could get out of control and lead to an “AI takeover.” Last month, speaking to NBC News, he urged action before it was too late. A Netflix documentary, The AI Doc, released last month, calls on viewers to press Congress to protect the public from the technology’s risks.

The founders’ public-facing work has coexisted with projects that proved harder to sustain. Former employees said CHT shut down an online course for technology workers seeking to build more humane systems after thousands enrolled, when Harris shifted his attention to AI risks. Work on materials to help educators address screen time also faded. One former employee said Harris repeatedly returned to the idea that the center should serve as a platform for his message.

That emphasis was not just a matter of communication style. Former staff said Harris and the team sometimes disagreed over whether CHT should focus on users, technology companies or lawmakers. Some saw the organization’s approach as too focused on influential people and not enough on conventional grassroots organizing. Shah’s defense is that broad public communication creates pressure on decision-makers, and that Harris can reach people across political lines.

What gets left behind

I think the central question is not whether Harris can attract an audience; the organization’s history suggests he can. It is whether attention, briefings and private coordination can substitute for the slower work of building pressure through legislatures and courts. CHT’s announcement chooses the first approach, but does not explain how it will measure whether that choice changes what companies do.

The tension is sharper because former employees point to results from the work now being dropped. In 2021, state attorneys general across the US began investigating Meta and later sued, alleging that the company misled teenagers and parents. This year, Meta agreed to pay billions of dollars to settle some claims; juries in separate cases, including one that heard testimony from Raskin, imposed additional penalties. Other social platforms also settled similar lawsuits this year.

A technology-sector nonprofit leader praised CHT’s less visible work in state legislatures and courts in an interview with WIRED earlier this year. He now calls the decision to abandon that work disappointing, arguing that change comes through social movements and democracy, not celebrities and private meetings.

CHT’s former social-media strategy consultant Gail Barnes sees the decision as a retreat from projects meant to turn Harris’s fame into sustained pressure on companies and change inside them. Whether that trade makes sense, she said, depends on what is holding back AI progress: people not hearing critics’ arguments, or those arguments failing to carry institutional influence. CHT is betting on the first explanation. Its new model puts more of the mission on the founders’ reach—and less on the institutional work that could test whether that reach has consequences.