According to OpenAI (@OpenAI), GPT-5 has been updated with input from over 170 mental health experts to significantly improve ChatGPT’s performance in sensitive conversations, reducing problematic responses by 65-80%. This update marks a concrete step for AI in supporting mental health use cases, enhancing trust and safety for both users and enterprise clients. The collaboration aims to address critical gaps in AI-driven support systems, offering new opportunities for businesses in the digital health, teletherapy, and mental wellness sectors to integrate more reliable AI-powered assistants. (Source: OpenAI, https://openai.com/index/strengthening-chatgpt-responses-in-sensitive-conversations/)

According to @AnthropicAI, the company is advancing research into the affective uses of its Claude AI model by leveraging privacy-preserving tools and collaborating with mental health experts such as @throughlinecare. This initiative focuses on equipping Claude to handle emotionally challenging conversations, signaling a significant step toward safe and responsible AI integration in mental health support. By prioritizing data privacy and consulting with clinical professionals, Anthropic aims to open new business opportunities in AI-powered digital therapy, customer support, and sensitive user engagement while maintaining high ethical standards (Source: AnthropicAI on Twitter, June 26, 2025).