Can we use AI to close the gap in women’s mental health support?

Here, balance looks at how the disparity in women’s mental health could utilise AI tools. But can this technology reduce gaps rooted in decades of inequity, without amplifying them?

In a world increasingly shaped by technology and the digital world, the promise of artificial intelligence (AI) to enhance health care is both extremely exciting and exceptionally contested. In no other industry is the promise clearer, whilst the stakes remain higher. Especially when we transfer AI to the topic of mental health.

Events like the Mindful Future Summit in London bring clinicians, researchers, and advocates together to spotlight both the promising future and risks of AI for women’s mental health. They ask the important question: Can technology reduce gaps rooted in decades of inequity, without amplifying them? In an era where caregivers take on emotional labour without support, AI can extend reach to those who may not be able to access these tools.

Women have historically faced distinct mental health challenges rooted in biological, social, and systemic inequality, and as digital tools snowball, questions about whether AI can close the gap in access, support and outcomes remain a heated topic.

The pressures shaping women’s mental health are intensifying at the exact moment technology is reshaping care. Women still carry the majority of unpaid emotional and caregiving labour, are more likely to work in insecure or underpaid roles, and experience higher exposure to gender-based violence and online harassment. All of which are proven risk factors for anxiety and depression.

At the same time, post-pandemic burnout, rising living costs, and the medicalisation of stress through wellness culture have left many women navigating these issues in an isolation epidemic. If AI is going to become a new gatekeeper of our care (a skilled and historically human job that requires layers of emotional intelligence) through chatbots, digital tools, and predictive systems, then whose issues it recognises, how it translates symptoms, and who it prioritises will matter a lot.

Why do we all need to understand the gap in women’s health?

Before AI even entered the picture, gender disparities in mental health were present. Research consistently shows that women are nearly twice as likely as men to be diagnosed with depression and anxiety disorders, including PTSD and eating disorders, across many populations and cultures worldwide. Women’s rates of depressive disorders have been reported at roughly 5.8% vs 3.5% for men over 12 months in large surveys.

Data from mental health foundations and health research underscores that young women (aged 16-24) are almost three times as likely as young men to experience common mental health issues. These disparities reflect social inequalities, such as gendered caregiving burdens, economic inequality, violence exposure, and cultural stigmas around seeking help, which historically have shaped women’s mental wellbeing at every stage of life.

Notably, caregiving roles themselves (including parenting, managing household stress, supporting elderly relatives) carry psychological strain that has been tied to heightened emotional volatility and stress in women. Large-scale studies show that emotional variability and mental health instability are higher in women who are caregivers, compared with other groups.

But, this is also not a suggestion of pitting men’s mental health against women’s, or proposing that one deserves attention at the expense of the other. Mental health inequity harms everyone. Closing a gap requires looking closely at where the gap actually is. And for the reality of many women, distress is shaped by structural issues that technology alone cannot see unless it is designed to. Women are more likely to be primary caregivers, more likely to live with chronic stress linked to low pay or unstable work, and more likely to experience trauma connected to gendered violence and reproductive health. Highlighting these experiences is not about exclusion of one versus another. By understanding how mental health manifests differently across social roles and life stages, intelligent tools can be trained to respond more accurately, more humanely, and more effectively, therefore improving the overall quality of care systems for everyone.

Is there promise for AI in our mental health systems? And how are we already advancing?

1. Accessibility and support tools

AI chatbots and digital platforms like Wysa, which uses conversational AI based on CBT (Cognitive Behavioural Therapy) and DBT (Dialetical Behavioural Therapy) techniques, are expanding access to therapeutic tools globally.

The platform Wysa is now present in over 100 countries and has been adopted by national health services, a strong case in showing how AI can scale support beyond traditional clinics.

Recent polling in the UK found that more than one in three adults have used AI chatbots for mental health or wellbeing support, especially among younger people (with usage peaking at 64% among 25-34 year-olds). These tools can reduce barriers like cost, travel, stigma, and NHS wait times, which can also be crucial for women who may struggle under traditional systems.

2. Helping our providers

Emerging research suggests that AI can help caregivers and clinicians deliver more consistent and empathetic care. For example, an AI-assisted therapy platform significantly reduced response times and improved empathic care delivery, compared with non-AI assistance, by supporting providers in structuring therapeutic responses.

This means care can be streamlined without becoming colder or too robotic. It allows clinicians to spend less time drafting responses and more time offering the human connection that patients actually need. In a hybrid response, we can utilise AI whilst still keeping our human interaction and emotional connection.

3. Detecting it early and personalising our responses

Systematic reviews of AI use in mental health report that tools like predictive modelling and chatbots can help with early detection and monitoring. We can then look to tailoring interventions and engaging users earlier than traditional pathways might allow, bypassing wait times and even human error in issues going undetected. Users even describe AI as a “digital companion” and “safe space” that can help suggest self-care activities or prompt reflection when symptoms shift quickly and unexpectedly.

But can AI actually close the women’s mental health gap?

1. There is still bias in AI models…

Research has shown that AI tools risk perpetuating historical bias. Commonly used AI models in health and social care sometimes downplay women’s physical and mental health needs compared to men’s, describing identical symptoms in women in less serious terms. This disparity can affect care decisions.

AI trained on male-dominant clinical data may misinterpret women’s symptoms or fail to recognise gender-specific presentations of conditions like depression or anxiety.

2. The ethics of it all

Women surveyed about AI used in mental health services emphasise the importance of understanding how AI works, how data is used, and of preserving the human-clinician connection. Concerns about misdiagnosis, data privacy, and loss of therapeutic connection were all brought up. Could it do more harm than good?

3. Human connection still matters!

Mental health is deeply relational and cannot fully replace human empathy. By trying to algorithmically replicate nuanced judgements, we may breed a culture of untrust when it comes to women seeking help. It should be emphasised that AI should help but not replace human care when used correctly.