Mila launches its first national hackathon focused on the safety of conversational AI in youth mental health contexts

In a significant move towards ensuring the safety of youth interacting with conversational AI, Mila – Quebec Artificial Intelligence Institute, in partnership with Bell Canada, Buzz HPC, and Kids Help Phone (KHP), has announced its inaugural national hackathon. Scheduled to run from March 16 to 23, 2026, the event aims to develop and implement technical safeguards that mitigate risks associated with conversational AI systems, particularly in contexts related to youth mental health.

The hackathon's primary objective is to create, test, and document practical measures that can enhance the safety of young users engaging with conversational AI. This initiative comes at a time when such tools are increasingly integrated into daily life, yet there remains a pressing concern regarding their potential to exacerbate psychological vulnerabilities when not designed or governed appropriately. By bringing together experts in artificial intelligence, data science, and cybersecurity, alongside mental health professionals, the event seeks to address these critical issues head-on.

The Role of Conversational AI in Youth Mental Health

Conversational AI tools are rapidly becoming a staple in various sectors, including mental health support. These systems are designed to provide assistance, guidance, and support to users, particularly young individuals who may be experiencing emotional distress. However, the reliability of these AI systems is paramount. According to KHP, its services were accessed over 3.7 million times in 2025, highlighting the demand for effective mental health support. Furthermore, 78% of young users reported sharing sensitive information with KHP that they had never disclosed to anyone else. This underscores the necessity for conversational AI to deliver accurate and safe responses, particularly in vulnerable situations.

The hackathon will emphasize applied AI safety methods, focusing on the development and evaluation of safeguards, prompt engineering techniques, and the utilization of synthetic data to test risky behaviors. These approaches aim to enhance the robustness of AI models, ensuring they can effectively recognize risks and redirect users to human support when necessary. The event will kick off with a public opening conference on March 16, 2026, which will be accessible both in person in Montreal and online, allowing broader participation.

Collaborative Efforts for a Safer Future

The collaborative nature of this hackathon is noteworthy. Mila, renowned for its contributions to AI research and development, has partnered with key organizations like Bell and KHP to leverage their expertise in technology and mental health. As articulated by KHP's Group Head and Executive Vice President, Jenny Yuen, there is a collective responsibility to ensure that conversational systems are designed with safety and rigor, especially when they engage with vulnerable populations. This partnership aims to blend human and AI capabilities, creating a safer generative AI experience for young users.

Mila's background as the world's largest academic AI research center specializing in deep learning positions it as a leader in this initiative. Founded by Professor Yoshua Bengio, the institute is dedicated to advancing scientific breakthroughs that ensure AI benefits society. Supported by the Canadian and Quebec governments, Mila continues to influence the global landscape of responsible AI development.

Bell, as Canada's largest communications company, plays a crucial role by providing the technological infrastructure necessary for the hackathon. Their expertise in advanced networks and AI-driven solutions will be instrumental in developing the safety mechanisms that are the focus of this event.

Why it matters

The implications of this hackathon extend beyond the immediate goals of developing safeguards for conversational AI. As AI systems become more prevalent in mental health contexts, the need for responsible design and implementation becomes increasingly critical. This initiative not only seeks to enhance the safety of young users but also sets a precedent for future developments in AI technology. By prioritizing safety and reliability, the event aims to foster a culture of accountability within the AI community, ensuring that innovations are aligned with ethical standards and societal needs.

For creators and technologists, this hackathon represents an opportunity to engage with pressing issues in AI safety. The focus on applied methods and collaboration with mental health professionals highlights the importance of interdisciplinary approaches in technology development. As AI continues to evolve, the lessons learned from this event could inform best practices and guidelines for creating responsible AI systems that prioritize user safety, particularly in sensitive areas like mental health support. The outcomes of the hackathon could potentially influence future policies and standards in the AI industry, making it a pivotal moment for all stakeholders involved.