Getting Started

Access to mental healthcare is limited worldwide, partly because there is a shortage of trained mental health professionals, especially in low-resource settings. However, according to analysis from the McKinsey Health Institute, closing this gap could result in more years of life for people around the world, as well as significant economic gains. Task-sharing models, which allow trained non-mental health professionals to deliver evidence-based mental health services, can be a powerful solution to help bridge this access gap for mental healthcare.

To maintain the quality and sustainability of these models, they require structured processes, robust supervision, and an intensive investment of talent, which can be challenging. Solutions that use AI can help address these challenges. By standardizing training and avoiding the need for a human to be involved at every phase of the process, AI can help mental health task-sharing programs effectively scale evidence-based interventions throughout communities, maintaining a high standard of psychological support. Designing systems with appropriate oversight and diverse perspectives can impact the lives of many more people worldwide. 

This field guide introduces a model by which AI could help mental health task-sharing programs scale by supporting specialists and nonspecialists performing their respective roles. It provides foundational knowledge, actionable strategies, and real-world examples to responsibly use AI to effectively scale their work.

The field guide covers several areas, including:

• an introduction to AI and large language models (LLMs)
• potential use cases of AI in mental health task-sharing programs and real-world examples
• a capability assessment to determine how ready a task-sharing program is to adopt AI technologies
• examples of technology, quality, safety, trust, and regulatory considerations, and governance that can help reinforce the financial and operational sustainability of the AI solutions
• additional resources for task-sharing programs looking to adopt AI

It is important to acknowledge that the application of AI in task-sharing models is new and only a few pilots have been conducted. Many of the ideas outlined in this field guide are theoretical and have not yet been widely tested in real-world settings. Rather than offering a prescriptive guide, this field guide aims to introduce possibilities and inspire mental health programs to explore AI’s role in expanding access to evidence-based interventions. 

We invite you to use this resource to inspire opportunities for how AI can be used as a mechanism to accelerate the impact of your programs. We acknowledge that some of the use cases and examples presented may require additional consideration when it comes to how they compare to the specifics of other countries or circumstances. Our aspiration is to collaborate with others iterating and building out this Field Guide out to align with the needs and realities of diverse contexts.

Therefore, this field guide is not intended to:

• Provide a step-by-step guide or technical toolkit for developing and implementing specific AI solutions
• Advocate for the effectiveness and/or feasibility of AI solutions in mental health task sharing programs
• Advocate for or provide examples of clinical uses of AI
• Analyze or document the clinical evidence of particular AI solutions or models

This field guide is designed for mental health programs that train and deploy individuals to provide support or care in community-based settings. These individuals include volunteers, peer support specialists, laypeople, and licensed clinicians, such as doctors and nurses. It is also intended to show that there are viable opportunities to better unlock the potential of task sharing that require funding.

This field guide was primarily developed in collaboration with task sharing programs and, therefore, uses them as the primary use-case example. These programs are specifically designed to address the global mental health workforce shortage by expanding access to evidence-based care, particularly in underserved communities, so they are a good starting point for exploring how AI can support in scaling mental health service delivery.

That said, many of the AI use cases and topics explored throughout the field guide may also be relevant across the broader spectrum of health skill-building programs. For example, they may be relevant for those that train peer support specialists and volunteers to provide support or upskill licensed professionals. The first few chapters provide more general background and guidance, while the last few dive deeper into a more specific and specialised set of steps and considerations for implementers of task-sharing programs.

Whether your program is just beginning to explore AI or is already piloting AI solutions, we hope this field guide provides insights to support your journey.

Although we encourage you to read the entire document, this field guide is structured to be flexible and modular, allowing mental health programs to use it based on their specific needs. You can:

• Explore use cases and real-world examples to identify solutions that could enhance your program.
• Dive into specific chapters for guidance on technical considerations, safety, and regulatory compliance, governance, and sustainability.
• Apply the frameworks and checklists to assess your program’s readiness to implement AI, ensure responsible AI use, and develop implementation plans.

The content of the field guide was developed and validated through:

• 10+ discovery interviews with experts from health task sharing programs and global institutions
• An in-person summit that featured panel discussions with 20+ experts from mental health programs, including task-sharing and crisis line/peer support programs, and global institutions
• 5+ real-life case studies from organizations using AI to improve training, supervision, and intervention quality in mental healthcare
• Review of 100+ peer-reviewed articles on mental health task sharing and the use of AI in healthcare and mental health
• Review of existing AI frameworks and guidelines in healthcare and mental health published by leading global organizations (including the WHO, the World Economic Forum, and the Coalition for Health AI)
• Input from 15+ mental health and AI experts to ensure the applicability and relevance of the content

This field guide was developed by Google for Health, Google.org, Grand Challenges Canada and McKinsey Health Institute, with contributions from experts at organizations such as Born This Way Foundation, CETA, Child Mind Institute, Children’s Hospital of Philadelphia, Fountain House, Friendship Bench, Georgia Tech, Global Mental Health Lab at Columbia University, Harvard T.H. Chan School of Public Health, Jack.org, Jed Foundation, Mental Health America, Mental Health for All Lab at Harvard Medical School, National Council for Mental Wellbeing, Partnership to End Addiction, Reflex AI, Stanford Medicine, StrongMinds, Substance Abuse and Mental Health Services Administration, The Trevor Project, UNICEF, Wellcome, World Health Organization, and 7Cups. Expert contributions represent inputs from individuals and do not imply endorsement from the organizations with which they are affiliated.

This field guide is a testament to the collaborative spirit of the mental health and AI communities, and we are grateful to be part of the collective effort to advance responsible AI use in mental health.