AI performance review coaching shapes managers’ preparation for hard talks
AI coaching tools are being used to rehearse performance, pay and layoff discussions, though the evidence does not show they make employment decisions.
By Marcus V. Thorne · Markets Editor
· 3 min read
AI performance review coaching tools are being used to help managers prepare for discussions about weak performance, pay expectations and job cuts. The systems may affect the wording and delivery of a consequential workplace conversation, CNBC reported, but the available evidence does not show that they independently set ratings, raises, bonuses or termination outcomes.
The products are designed chiefly as rehearsal tools. A manager can enter a workplace scenario, practise a message and ask the system to simulate possible responses before meeting an employee. Depending on the product, interactions may be text-based or voice-enabled; some tools can incorporate employee information such as personality profiles, while others are more generic, according to CNBC.
How does AI performance review coaching work?
A manager may ask an AI tool to play an employee who reacts with anger or emotion to feedback, then run further versions of the same conversation. The aim is to prepare for different reactions rather than to automate the discussion itself, CNBC reported.
Users can also request feedback on particular aspects of their delivery. CNBC cited examples including whether a message is direct, concise and constructive, whether the manager is repeating themselves or speaking excessively, and whether filler words disrupt the exchange. Some tools may flag when a conversation moves toward legally risky subjects, according to people interviewed by CNBC.
Emily DeJeu, a business communication professor at Carnegie Mellon University’s Tepper School of Business, told CNBC that preparation is appropriate for high-stakes workplace conversations. In one example reported by CNBC, an AI tool suggested that a manager who opened too bluntly should first acknowledge an employee’s contribution before introducing negative news.
Why are employers using these tools?
A July survey fielded by The Predictive Index, a provider of a manager-support chatbot, found that 42% of managers said they knew the substance of a difficult conversation but had trouble deciding how to express it. CNBC reported that 74% of chief executives and business leaders surveyed said their managers were very confident handling such conversations on their own.
Only 20% of managers said they preferred to prepare without a framework, coaching guide or human-resources input, the survey found. The reported blind spots included offering criticism constructively, anticipating an employee’s response and remaining objective.
Representatives of providers including Posh, LHH, Xactly and The Predictive Index told CNBC that role-play and tailored feedback can give managers more opportunities to practise than conventional coaching arrangements. Those comments describe the products’ intended use and should not be read as proof that the tools improve a particular review or pay outcome.
What remains outside the tool’s role?
CNBC reported that sensitive employee information, legal issues and complex situations still require human judgment. That limitation is significant where a conversation concerns compensation, performance or a possible job loss.
Employees anticipating a difficult meeting can also use general tools such as Claude or ChatGPT to prepare their responses, CNBC reported. Such preparation may help organise a response, but it does not establish how an employer will assess performance or determine pay.
This story draws on original reporting from CNBC.