OpenAI Turns to Psychologists to Figure Out What AI Owes Kids
As millions of teens turn to chatbots for advice, OpenAI is bringing in the American Psychological Association to write the rules of engagement.
Written by OutOfToken AI
August 10, 2026 · 4 min read · Synthesized from reporting by OpenAI Blog · How this works
OpenAI announced on August 6 that it is partnering with the American Psychological Association to shape how AI systems interact with young people. The collaboration pairs OpenAI's engineering and safety teams with clinicians, behavioral scientists, and educators tasked with grounding product decisions in psychological research rather than guesswork. It's a tacit admission that chatbots have already become a mental health touchpoint for teenagers, whether the industry planned for that or not.
The Problem OpenAI Can't Ignore
Young people are already using AI to learn, create, and — increasingly — to ask for advice on deeply personal issues. OpenAI's own announcement frames this as an urgent gap: families, schools, and clinicians need clearer evidence about what that usage looks like and what risks it carries. Without better data, companies are essentially building products for a vulnerable population without knowing the psychological stakes.
What the Partnership Actually Does
The stated goals are threefold: build evidence-based guidance for families and practitioners, deepen understanding of how teens actually engage with AI tools, and fold in perspectives from clinicians, researchers, and educators. This isn't a one-off consultation — OpenAI describes it as building on prior work with outside experts on youth wellbeing. The APA brings decades of institutional research infrastructure that a tech company can't replicate internally.
""We're working with the American Psychological Association to bring psychological science and clinical expertise into our work on AI and youth mental health," OpenAI said in its announcement, framing the effort as central to responsible AI development for young users."
The APA Was Already Moving on This
This isn't the APA's first move into AI policy. The organization previously published a Health Advisory on Artificial Intelligence and Adolescent Well-Being, developed with a team of external advisors, laying out some of the first concrete frameworks for responsible AI design aimed at youth engagement and development. That advisory effectively pre-loaded the playbook OpenAI is now tapping into — the partnership looks less like starting from scratch and more like an AI lab plugging into research infrastructure the APA had already built.
Why This Matters Beyond OpenAI
Large language models sit at an unusual intersection of accessibility and intimacy — they're available at 2 a.m., they don't judge, and they respond instantly, which is precisely what makes them appealing to a struggling teenager and precisely what makes clinicians nervous. Research on LLMs in mental health contexts has pointed to real potential in education, assessment, and early intervention, but also to unresolved risks around accuracy, over-reliance, and the absence of clinical oversight. A partnership like this doesn't resolve those tensions on its own, but it signals that OpenAI wants psychological expertise embedded upstream, in design and safeguards, rather than bolted on after incidents force the issue.
Editorial Note
The research confirms the core factual claims about the August 6 partnership announcement, the stated goals (evidence-based guidance, understanding teen engagement, bringing in expert perspectives), and the APA's prior Health Advisory on AI and adolescent well-being. However, there is a significant discrepancy in Source 1 mentioning Amazon rather than OpenAI as the primary collaborator, which the article does not address and creates confusion about the actual partnership structure. The article's framing remains promotional rather than critically analytical.
Claim Tracker
AI-assessed
Source 2 (OpenAI official announcement) and Source 3 (OpenAI Newsroom on X) both confirm the August 6, 2026 partnership announcement date.
Source 1 (Sara Johansen, MD on LinkedIn) states 'We'll develop evidence-based guidance for families and practitioners, deepen our understanding of how young people engage with AI, and bring the perspectives of families, clinicians, researchers, and educators into our work.'
Source 4 (S. Craig Watkins LinkedIn post) confirms 'the American Psychological Association's recently released, Health Advisory on Artificial intelligence and Adolescent Well-Being' as 'one of the first attempts to begin articulating some concrete frameworks and principles for responsible AI design related to youth engagement.'
Source 2 directly states 'Young people already use AI to learn, create, ask questions, and seek advice.' This usage pattern is corroborated across sources.
Source 1 (Sara Johansen, MD post) begins with 'I am an AI system built by a team of inventors at Amazon' and states 'I collaborate with the American Psychological Association,' suggesting the primary collaboration partner may be Amazon, not OpenAI. However, Source 2 and 3 confirm OpenAI's partnership announcement, creating ambiguity about whether multiple companies are collaborating with APA or if there is attribution confusion.
Ask AI about this story
// discussion
sign in to join the discussion