A lot of management comes down to conversations that are hard to get right. A manager might need to address a performance issue, discuss an employee’s career concerns, or help someone navigate organisational change. Each situation calls for a different approach, and finding the right words isn’t always straightforward.
But even experienced managers can struggle with how to start these conversations, how direct to be, or what to do when the employee reacts differently than expected.
Many are now turning to AI for help.
Findings from The Predictive Index’s AI and People Leadership Report revealed that three in four managers have used public AI at least once to prepare for a people conversation. That gives them a fast source of guidance when they are unsure how to handle a difficult moment. The problem is that a general-purpose tool may know plenty about giving feedback, but nothing about the employee receiving it.
The new management sounding board
Managers facing difficult conversations may not have time to consult an HR partner or revisit formal training. AI offers an accessible way to prepare, helping them organise their thoughts and consider how a conversation might unfold.
It also gives managers a private space to work through their concerns before involving someone else. They can explore different approaches, refine their thinking and build confidence before having a conversation with an employee.
But quick advice is not the same as preparedness. In the same PI report, 74% of executives said their managers were very confident handling difficult conversations on their own. Asked separately, only 57% of managers said they entered those conversations fully prepared. These numbers point to a disconnect between leadership’s confidence in managers and how prepared managers actually feel.
Good advice still needs to fit the person
Knowing what to say is only part of a difficult conversation. Managers also have to consider how their message will be received.
The same feedback can land differently depending on the employee. Some prefer a direct approach with clear next steps. Others may need more context or time to process the information before discussing what comes next.
Generic AI tools have limited insight into these individual preferences. They can help managers structure a conversation, but the advice may not account for how a particular employee communicates or responds to feedback. Managers still need to decide how best to approach the person in front of them.
HR can’t prepare managers for every moment
Formal manager development still has an important role to play. Training can build core skills around feedback, coaching, performance management, and change.
But many managers are still stepping into people leadership without enough preparation. The American Management Association’s Talent Development Research surveyed more than 1,000 professionals and found that 51% entered people management without formal leadership training. Another 66% said they do not receive frequent support from their own managers.
Even with this foundation, managers will encounter situations that require additional support. HR teams also cannot be available whenever a difficult conversation arises.
HR can help bridge this gap by giving managers access to trusted resources when challenges arise. This could include practical guidance on handling sensitive conversations, along with behavioural insights that help managers understand the people involved. AI can make these resources easier to access, giving managers a starting point they can adapt to the employee and situation.
Bring behavioural intelligence into the moment
Organisations can draw on behavioural science to give managers a clearer understanding of how their employees work and communicate. The Predictive Index has spent decades studying workplace behaviour, providing science-backed insights into what drives individuals and how they prefer to interact with others.
These insights can help managers consider how an employee prefers to receive information and respond to feedback. Understanding those preferences gives managers a stronger foundation for approaching the conversation.
Consider a manager preparing to deliver difficult feedback. A general-purpose AI tool might recommend a direct conversation with clear next steps. Yet behavioural insights could suggest that the employee prefers time to process information before responding or works best when given the autonomy to explore solutions independently.
Armed with this understanding, the manager could use AI to explore different ways of framing the conversation. They might decide to share key points in advance or leave more room for the employee to contribute to the next steps.
Give managers context, not a script
AI and behavioural insights can help managers prepare for difficult conversations, but they should never replace human judgement. Managers must still assess the circumstances, question recommendations and take responsibility for their decisions.
Behavioural profiles can offer valuable context about an employee’s preferences, but they should not be treated as definitive predictions of how someone will respond.
HR can establish guidelines for responsible AI use, including how employee information should be handled and when managers must exercise their own judgement. These safeguards can help managers use AI and behavioural insights confidently while keeping the focus on the employee involved.
Explore The Predictive Index’s The Copy-Paste Problem to learn how behavioural intelligence can help managers make better use of AI guidance.
A lot of management comes down to conversations that are hard to get right. A manager might need to address a performance issue, discuss an employee's career concerns, or help someone navigate organisational change. Each situation calls for a different approach, and finding the right words isn't always straightforward.
But even experienced managers can struggle with how to start these conversations, how direct to be, or what to do when the employee reacts differently than expected.
Many are now turning to AI for help.
