Artificial intelligence is increasingly used in the workplace as an employee’s mental and emotional companion: chatbots that listen, offer coping strategies and track mood over time. This analysis from BTUAI examines what these tools do, how well they work, where the risks and the surveillance line fall, how regulation is changing, and what this shift means for Georgia.
In short: key takeaways
- Most credible sources agree that AI is a first, supporting layer – not a replacement for a psychologist.
- For mild-to-moderate, short-term symptoms the effect is real, but its durability over time is not confirmed.
- In a workplace context the sharpest risk is surveillance: the employer both provides the tool and benefits from its data.
- Regulation is moving fast: U.S. states and the EU already regulate “companion chatbots”.
How is AI used as an employee’s mental companion?
A purpose-built “clinical” AI companion operates within a defined scope. It offers psychoeducation for mild distress – coping strategies and self-help resources – as an always-available first layer, and it tracks mood, nudging a person toward more substantive support before a crisis. A separate “passive” mode observes not emotion but work rhythm.
Does it work? What does the evidence say?
The picture is mixed. Controlled trials showed a significant reduction in depressive symptoms, and real-world data revealed a dose-response relationship: the more often people used the tool, the more their condition improved. However, the effect is short-lived and fades in about three months, evidence in non-English languages is thin, and much of the research is funded by the vendors themselves.
What are the risks of an AI companion at work?
The main risks are emotional dependency, substituting professional care with an algorithm’s generic advice, and weak safeguards in a crisis. In a workplace context, surveillance is a distinct concern: “passive” monitoring easily shades into tracking. The International Labour Organization (ILO) calls for new rights — to an explanation of AI decisions, to human review, and to data protection.
How are AI companions regulated?
The law is catching up with the technology. Several U.S. states have passed laws on “companion chatbots”: mandatory transparency, self-harm prevention protocols and extra safeguards for minors. For employers this means stricter due diligence at procurement: the tool’s precise scope, a responsible clinician, a full crisis-escalation path and data-protection compliance.
What does this mean for Georgia?
In Georgia, workplace mental health is weakly institutionalised and professional capacity is thin, which makes a scalable “first layer” potentially valuable. At the same time, Georgian employees increasingly turn to general-purpose models for informal emotional support. The key difference is that the human “second layer” – accessible clinicians – is thin here, which easily turns the companion into a de facto replacement.
Where should a Georgian employer start?
The leverage is a scalable first layer where professional capacity is thin. But that benefit depends on three conditions: the tool must be well-built and validated in the Georgian language, there must be a clear path to hand a person over to human help, and there need to be clear data-protection rules. Because Georgia does not yet have comparable regulation, the responsibility rests with the employer.
This analysis was prepared by the academic team of the Business and Technology University (BTU) together with BTUAI and draws on publicly available industry reports and research and is analytical in nature. The full report, with charts, is attached as a separate file.
Citation: Business and Technology University (BTU) & BTUAI. (2026). Artificial Intelligence as an Employee’s Mental Companion. Tbilisi, Georgia.