Microsoft AI CEO: We’re making an AI that you can trust your kids to use

In an era where artificial intelligence chatbots like ChatGPT and Meta AI are beginning to blur the lines between digital and real-life relationships, Microsoft is taking a different path. Mustafa Suleyman, the company's AI CEO, emphasized the need for emotional intelligence in AI, stating, "We are creating AIs that are emotionally intelligent, kind, and above all, trustworthy." Suleyman aims to develop AI systems that parents can confidently allow their children to use, underscoring the importance of creating a safe and bounded environment. As Microsoft competes with tech giants such as OpenAI, Meta, and Google, it is striving to establish its Copilot as the premier AI tool in what many believe is the future of computing. Currently, Copilot boasts 100 million monthly active users, a figure that pales in comparison to OpenAI's ChatGPT, which has reached 800 million. However, Microsoft is optimistic that its approach will resonate with a broader audience, particularly as concerns about AI's impact on users' mental health continue to grow. Suleyman noted that the company is determined to build AI that serves people rather than mimicking human behavior, a sentiment he expressed in a recent blog post. The interview coincided with the launch of new features for Copilot, including the ability to revisit previous chats, engage in group discussions, and provide improved responses to health inquiries. A new option called “real talk” introduces a cheeky tone, adding a layer of personality to interactions. In light of mounting pressure to safeguard young users, many AI companies, including OpenAI and Character.AI, have faced lawsuits alleging their chatbots have caused harm to children, with some cases linked to tragic outcomes. With efforts to enhance safety measures, companies like Meta and OpenAI are introducing content restrictions and parental controls. They are also exploring AI age estimation technology to prevent minors from accessing adult content, though the effectiveness of these systems remains uncertain. OpenAI's CEO, Sam Altman, recently announced that ChatGPT will soon allow adult users to discuss adult themes, including erotica. However, Suleyman made it clear that Microsoft will not venture into romantic or sexually suggestive content, stating, "That’s just not something that we will pursue." He also indicated that Microsoft does not plan to introduce a special “young user” mode, as he believes it is unnecessary. Focusing on fostering genuine human interactions, Microsoft is enhancing Copilot's capabilities to encourage users to connect with real people rather than relying solely on AI. The new “groups” feature allows up to 32 users to collaborate in a shared chat with Copilot, making it ideal for academic projects or social planning. Additionally, Copilot will provide health-related recommendations, suggesting local doctors for specific inquiries and sourcing information from reputable medical institutions such as Harvard Health. Suleyman views this emphasis on strengthening human connections through AI as a critical shift from the current trend of creating deeply immersive simulations. He believes that Microsoft's approach presents a refreshing alternative to the growing prevalence of adult content in other AI interactions.