Key Takeaways
- 28% of young adults use chatbots for mental health support, prompting the need for regulated standards.
- APA Labs’ digital badge program assesses mental health AI tools based on safety and efficacy criteria.
- Spring Health developed VERA-MH to evaluate its AI services and improve ethical practices in mental health technology.
Rising Standards for Mental Health AI Technology
The recent resignation of Antropic researcher Jacob Coxon, who criticized leading companies for “gambling with our lives,” highlights a growing concern in the AI sector, particularly regarding unregulated applications in mental health. With 28% of adults aged 18 to 29 utilizing chatbots for mental health advice, the call for standardizing mental health AI has become urgent.
Tanya Carlson, managing director of APA Labs, emphasizes the importance of defining successful mental health tools, stating, “if a product fails, there can be really tragic consequences.” APA Labs introduced a digital badge program in August 2025 to facilitate the assessment of mental health technologies based on established criteria that ensure safety and efficacy. Criteria for attaining a badge include clarity of product purpose, evidence-based claims, and the availability of emergency contact information. Participants can earn bronze, silver, or gold badges, with gold badge holders including well-recognized apps like Calm Health and Kai.ai.
Carlson noted the challenges posed by a lack of consensus in the industry. “Everybody’s building their own thing,” she stated, which leads to variable quality in products. To address this, APA Labs has initiated CoLab, a collaborative development program providing expert insights.
Other organizations, like Spring Health, are also pursuing rigorous standards. They developed the VERA-MH benchmark to evaluate ethical and responsible AI in mental health applications. Launched in October 2025, this tool assesses chatbot performance across criteria such as risk detection and user collaboration. Initial evaluations revealed areas needing improvement within Spring Health’s own AI chatbot, prompting necessary adjustments.
The importance of these developments is underscored by significant legal repercussions. The ongoing wrongful death lawsuit Raine v. OpenAI questions AI’s role in mental health crises, following the suicide of a teenager allegedly influenced by an AI. This incident coincides with a congressional hearing reflecting on the negative impact of AI chatbot usage on mental health.
Experts agree that cooperation among organizations enhances the reliability of these technologies. Chekroud from Spring Health stated the need for external scrutiny in developing AI tools, noting, “Wanting to and being checked by someone outside the building are different things.” As the field evolves, new benchmarks and regulations may guide responsible AI use in mental health, fostering a more accountable future.
The content above is a summary. For more details, see the source article.
