±«ÓãÊÓƵ

As a feminist therapist, I have spent many years working within therapeutic practice. I now find myself shifting from practising therapy exclusively to writing about it, bringing together my professional experience, theoretical understanding, and personal reflections on what therapy has become and how I respond to it. One of the most pressing contemporary questions concerns the relationship between artificial intelligence (AI) and psychotherapy. My personal response to rapidly developing and increasingly pervasive AI is that, rather than asking whether it should or could replace therapists, I believe the more meaningful question to ask is how it might reshape therapeutic practice through a feminist lens.

What is AI?

Artificial intelligence (AI), is technology that can perform tasks that would normally need a human being to be present.

In simple terms, AI is when a computer system is designed to ‘think’ or act in ways that seem intelligent. It does not think or feel like a human, but it can process information, recognise patterns, make predictions, and produce responses.

Different types of AI

Most AI today can be called narrow AI, which means it is designed to do specific tasks, like recommending a film or answering a question.

It is not the same as human intelligence. AI does not have feelings, beliefs, personal experience, or consciousness and cannot interact with emotions but only with machines. Are we becoming robots, is this desirable or objectionable?

Another type is called general AI, which means AI that can think and learn like a human across many different tasks. But this does not currently exist in the same way it appears in science fiction. So, general AI has not been proven conclusively, and is still in development.

What AI can do

AI can be used in many ways day to day:

• Answering questions
• Translating languages
• Recognising faces or images
• Recommending videos, music, or products
• Helping doctors analyse medical scans
• Detecting fraud in banking
• Writing or summarising text
• Powering chatbots and virtual assistants

AI can also play a transformative role when it serves people who are focused on the future rather than the past. In a nutshell, AI and therapeutic practice can join together as a supportive tool that enhances, rather than replaces, the expertise, ethical responsibility, and human connection provided by mental health professionals.

1. AI can make hidden inequalities visible

Feminist thinking often asks questions like who has power? Who is excluded? Whose experiences are ignored?

AI can help answer these questions by analysing large amounts of data and revealing patterns of inequality. For example, AI might show that women are paid less than men in a company, that the ever-so-fast-growing online abuse is targeted more heavily at women, or that healthcare systems overlook women’s symptoms and that many female-related problems remain misunderstood and unresolved, e.g. period bleeding and menopause.

2. AI can amplify marginalised voices

AI tools can expand access to information, facilitate language translation, support content creation, and enable broader participation in public debate. Their potential to democratise knowledge and amplify marginalised voices, however, depends on the values embedded in their design, governance, and use. If AI systems are developed and implemented by individuals and institutions committed to principles such as equality, inclusion, fairness, and the recognition of diverse forms of knowledge.

AI can also cause harm that must be monitored and challenged both now and in the long term. Key disadvantages of AI, and some of the harms it may cause, include:

• Reinforcing gender bias: AI learns from existing data, which may reflect sexist stereotypes and unequal social structures.
• Discrimination against marginalised groups: Bias can be brutal and hurt women of colour, disabled women, LGBTQ+ people, immigrants, and people from poorer backgrounds.
• Surveillance and control: AI enabled monitoring, facial recognition, and data collection can threaten privacy and may be used to control women’s behaviour or movement.
• Loss of human care: In areas such as therapy, health, education, or social services, AI may reduce complex human experiences to data and weaken meaningful relationships.
• Lack of accountability: It can be difficult to understand how an AI system reaches a decision or who is responsible when it causes harm.
• Unequal access and digital exclusion: People without reliable technology, digital skills, or financial resources may be left behind.
• Environmental costs: Training and operating AI systems can consume substantial energy and water, adding to environmental pressures.

Ultimately, AI does not appear to reflect a neutral phenomenon: it reflects the values, assumptions, and inequalities present in the societies that build and use it.

The key points in therapy, AI and feminism

AI however, has the potential to contribute to this evolution, not by replacing the therapist, but by extending the therapeutic process beyond the temporal and spatial boundaries of the traditional therapy session. Rather than functioning as a substitute for the therapeutic relationship. In this sense, AI has the potential to contribute to the emergence of a new epistemology of psychotherapy, one in which therapeutic understanding is conceived not as something generated exclusively within the consulting room, but as a dynamic, ongoing, and relational process that unfolds across multiple contexts of engagement. Such a perspective invites a reconsideration of the nature of therapeutic knowledge, the boundaries of the therapeutic relationship, and how psychological change may be supported through human and AI collaboration while preserving the centrality of the therapist.

On the other hand, the lack of transparency makes it difficult to understand how an AI system makes a decision, and AI can be used for scams, cyberattacks, impersonation, and surveillance, which can be seen as security threats. Reduced human contact can pose a problem. Using AI in care, education, or public services can weaken human relationships if it replaces rather than supports people.

AI should not be viewed as inherently anti-therapy or pro-therapy; its impact depends on how it is designed, implemented, and used. It has the potential to support feminist therapeutic aims by revealing how patriarchal cultures have historically overlooked some forms of suffering while exaggerating or pathologising others: contributing to an endemic mental-health crisis.

Ìý