OpenAI’s KOSA Endorsement Draws Fire Over Regulatory Motives
OpenAI publicly endorsed the Kids Online Safety Act on May 13, 2026, igniting regulatory capture concerns.
Why it matters: OpenAI’s prominent support of KOSA spotlights the uneasy balance between child safety regulation and AI industry influence on US policymaking. As AI companies face growing legal challenges, their political engagement shapes the contours of future compliance—and competition.
- OpenAI’s endorsement of KOSA was announced on May 13, 2026.
- KOSA requires platforms to let minors opt out of algorithmic recommendations and addictive features.
- Critics say OpenAI’s move favors regulatory capture, benefiting industry incumbents.
- Digital rights groups warn KOSA could lead to censorship without robust child safety gains.
OpenAI’s public endorsement of the bipartisan Kids Online Safety Act (KOSA) has intensified scrutiny of how tech giants are shaping AI regulation. The endorsement, announced on May 13, 2026 by Chief Global Affairs Officer Chris Lehane, frames the legislation as a corrective to social media-era missteps: “We can't repeat the mistakes made during the rise of social media, when stronger safeguards for teens weren't put in place until the platforms were already deeply embedded in young people's lives.”
- KOSA, first proposed in 2022 and passed by the US Senate in 2024, aims to shield minors by requiring platforms to enable opt-outs from "addictive" features and algorithmic recommendations for users under 17.
- OpenAI's endorsement comes as the company faces lawsuits alleging harmful incidents involving minors, such as a wrongful death claim regarding ChatGPT.
- Advocacy and digital rights groups, including the Electronic Frontier Foundation, voice strong objections to KOSA, contending it risks functionally censoring online content without meaningful safety improvements for minors.
- Critics, including Techdirt and Fairplay’s Josh Golin, argue that OpenAI’s high-profile support is a "strategic move" to steer regulatory structures in ways that safeguard established companies at the expense of emerging rivals. Golin said, “OpenAI's endorsement of the bill also does nothing to change the myriad threats that OpenAI itself poses to children's health and safety through its dangerous products.”
The debate over KOSA reflects broader questions about who sets the standards for protecting children online—and who stands to gain from the resulting regulatory landscape.
By the numbers:
- May 13, 2026 — Date of OpenAI’s KOSA endorsement
- 2022 — Year KOSA was first introduced
- 2024 — Senate passage of KOSA
Yes, but: Details on how OpenAI will implement KOSA safeguards in its products remain unclear.