The recent letter published in the Financial Times—arguing that AI chatbots are not about to replace the healing that therapy offers—touches on a profound and growing fault line in how society approaches emotional wellbeing. (Financial Times)
At first glance it may seem like an expert’s pushback against tech hype. But beneath the surface lies a clash between business incentives, the realities of human psychology, and the economics of care provision. This is not just a debate about technology — it’s a debate about who gets help, what counts as help, and who profits from it.
Why This Conversation Matters Now
Two trends collide here:
- Mental health demand far outstrips supply. Across developed economies — the UK, US, EU — anxiety, depression and stress-related disorders have surged in recent years, especially among the young and working age. Professional services are overwhelmed and waitlists are long.
- AI chatbots are cheap, scalable and marketable. Investors and tech firms see emotional support as fertile ground. Unlike other healthcare verticals, mental health lacks the stringent regulation of pharmaceuticals or surgery — potentially opening a lucrative frontier for AI tools.
In principle, an AI that can converse comfortably with someone in distress sounds appealing. But ease of access does not equate to therapeutic effect.
Who Benefits — and Who Is Shortchanged
The Winners
Tech companies and investors AI chatbots are low-cost to deploy and can be monetised through subscriptions, data insights, or corporate licensing. They promise engagement at scale with minimal human labor — a potent value proposition for shareholders.
Users looking for immediate interaction For many individuals facing mild emotional discomfort or curiosity, a chatbot offers instant responses without stigma or scheduling.
But caution is warranted here: access does not guarantee effective treatment.
Those Losing Out
People with serious or complex needs Professional therapy is not merely conversational. It involves assessment, diagnosis, emotional attunement, and tailored intervention — elements currently beyond chatbots. The letter’s author, a seasoned psychotherapist, stresses that healing in therapy cannot be replicated by scripted algorithms. (Financial Times)
Therapists and mental health professionals There’s a risk that the rise of AI tools becomes a cover for underfunded systems. Rather than expanding real therapeutic capacity, policymakers and insurers might opt for cheaper, superficial substitutes, undermining the profession and the quality of care.
Healthcare systems and employers Incentivising unregulated digital tools as stand-ins for clinical care could lead to worse long-term outcomes — more chronic issues, repeated crises, and higher costs downstream.
Business and Industry Impacts
A New Tiered Mental Health Market
What’s emerging is a two-tier model:
- Low-touch, AI-mediated self-help products (cheap, scalable, minimal clinical oversight)
- High-touch, professional therapeutic services (expensive, limited supply)
This bifurcation mirrors broader trends in healthcare: commoditisation of routine touchpoints and premium pricing for human expertise. Tech firms will market chatbots as first-line solutions, capturing huge user bases. Meanwhile, therapists may be relegated to complex cases that AI cannot handle — but those are precisely the cases that strain health budgets and require long-term commitment.
The Hidden Implication: Data, Ethics and Dependency
There’s a quiet, less-visible aspect to this shift: data. Every interaction with a mental health chatbot generates behavioural signals that can be analysed — sometimes without clear consent or regulation. In other words, emotional data becomes a commercial asset before it becomes a clinical record.
This raises serious ethical questions:
- Who owns and benefits from emotional engagement data?
- Can such data be used for targeted advertising or influence?
- What happens when users become dependent on automated empathy, rather than human connection?
These are not abstract concerns — they are the real costs behind a rapid rollout of AI emotional tools.
Long-Term Effects: Not Just Therapy, But Society
If policymakers, insurers, and employers increasingly endorse AI chatbots as equivalents of professional care, the ramifications could extend far beyond individual health:
1. Devaluation of Human Care Work
Therapists, counsellors, social workers and community supports may find their roles undervalued, underfunded and marginalised — even as demand for their expertise climbs.
2. Shifting Norms Around Help-Seeking
People may come to see automated interaction as the default for emotional distress — similar to how generic wellness apps have substituted for clinical treatment in other domains.
3. Regulatory and Liability Challenges
There is currently no robust framework governing the use of AI in mental health — no standardised outcomes measures, no clear accountability for harm, and no licensing requirements. This regulatory gap will attract innovators and opportunists alike.
Why Human Healing Resists Automation
The FT letter’s core insight is that healing — real, lasting psychological change — is not a transactional conversation. It is relational, nuanced, interpretative, and deeply human. (Financial Times)
AI can simulate empathy — but simulation is not synthesis. It can offer reassurance, pattern recognition, and cognitive reframing. What it cannot do is engage with the whole person in context: history, relationships, trauma, resilience, identity.
That is why chatbots may serve as adjuncts — tools within a broader system — rather than replacements.
The Real Debate Ahead
The question for society is not whether AI chatbots will be used — they already are. The real debate is:
- Will we let commercial incentives dictate the standard of mental health care?
- Can we integrate AI thoughtfully without eroding clinical quality?
- Who ensures ethical use of deeply personal data generated through these tools?
What’s at stake is not only the future of therapy, but the economics of emotional wellbeing and the social contract of care. If we get this wrong, we won’t just mislabel a technology — we will reshape how society understands healing itself.