What happened
CNBC reports that tools including Obi, Posh and Penny let managers role-play difficult workplace conversations, receive feedback on tone and wording, and practice different employee reactions. The report says managers should avoid entering identifiable or confidential employee information and continue involving human supervisors or HR.
CNBC reports that several companies are offering AI tools designed to help managers prepare for difficult conversations about performance, compensation, attendance or job loss. Examples named in the report include The Predictive Index’s text-based Obi, Posh’s AI role-play training tool and Xactly’s text- and voice-based Penny. Reported features include simulated employee reactions, feedback on tone and clarity, tailored scripts and warnings about potentially risky topics.
According to CNBC, managers can ask the systems to simulate employees who are angry, emotional or sensitive to criticism, then repeat the exercise under different conditions. The article also reports that managers may use general behavioral descriptions rather than names or identifiable personnel details. Access terms, pricing, availability and independent testing of these products are not established in the source.
Why it matters
AI coaching could give managers more accessible opportunities to rehearse high-stakes conversations, especially when human coaching is unavailable. But these systems do not replace judgment, legal review or empathy. The practical value depends on the quality of the feedback, the accuracy of the scenario, and the employer’s controls over sensitive workplace data.
CNBC cites a July Predictive Index survey in which 42% of managers said they know what they want to say in difficult conversations but struggle with how to say it. The report says the tools are intended to provide practice and feedback rather than automate the people-management decision itself. That could be useful for managers who have limited access to coaches or HR support.
The risks are substantial because performance discussions can contain sensitive personal and employment information. CNBC reports that human-resources professionals advised avoiding identifiable details in public AI tools and checking enterprise systems’ access, retention and model-training policies. The article also emphasizes that human supervisors or HR should remain involved, particularly in complex or legally sensitive cases. The survey findings and vendor claims were not independently verified in the report.
What to watch next
Watch whether employers adopt these tools with clear privacy, retention and review policies, and whether vendors provide evidence that their feedback improves outcomes. The source does not independently establish the tools’ effectiveness, pricing, availability or data practices.
Employers should clarify which AI systems managers may use, whether conversations are visible to administrators or other employees, how long data is retained, and whether prompts or outputs can be used to train models. Those controls will determine whether role-play assistance creates manageable risk or exposes confidential personnel information.
The source reports that AI systems can be overly agreeable, which may limit the usefulness of their criticism. Further reporting would need to establish whether these tools produce reliable coaching across different workplace contexts and whether their use improves communication or employee outcomes. No independent benchmark, controlled study, user results or vendor pricing is provided here.