According to PI’s 2026 AI and People Leadership Report, 67% of leaders say HR always or often coaches managers before a tough feedback conversation or conflict. HR support is the default experience for most managers heading into a hard conversation — at least according to leadership.
Managers, separately, were asked how often HR is available to help before more challenging conversations. Forty-three percent said HR is always available to help before a hard conversation, and 41% felt the support only shows up sometimes. Split almost evenly down the middle, reliable access to HR is closer to a coin flip than leaders realize.
When researchers asked managers what would help them most before a difficult conversation, only 11% chose time with HR. A general framework or checklist ranked highest, followed closely by coaching built around the specific employee they were about to sit down with.
HR can’t realistically provide individual coaching before every conversation happening across a company, so its greatest leverage comes from giving managers guidance they can apply the moment they need it. Managers want HR’s judgment built into something they can reach on their own, in the moment, without waiting on a person who can’t always be there.
What do managers need to feel more prepared and supported?
Managers need preparation built around the specific person they’re about to sit down with, not a generic approach applied the same way to everyone.
Our report also found that 42% of managers believe their last difficult conversation fell short because they knew what they wanted to say but struggled with how to say it. Another 26% couldn’t predict how the other person would respond, and 20% could plan the conversation but couldn’t adapt once it started going differently than expected.
None of this suggests HR is handing managers generic advice when they do ask for help. HR simply can’t sit in on every conversation happening across a company, and that limit shows up in how prepared a manager feels walking in. Without HR in the room, a manager is left guessing how a specific person will respond, and that uncertainty shapes the kind of guidance they end up relying on. A few common examples:
- Rehearsing one version of a feedback conversation, then improvising entirely when an employee reacts with defensiveness instead of the calm response they’d prepared for.
- Delivering the same framework to every employee, without adjusting for someone who needs more context before receiving criticism.
- Avoiding a harder truth altogether, unsure whether direct feedback will land as constructive or combative with that particular person.
Seventy-six percent of managers say guidance tailored to the employee they’re meeting with would help them prepare. Insight into how a specific person communicates and receives feedback gives managers a clearer sense of how the conversation is likely to unfold before they’re in it.
How behavioral intelligence makes HR’s judgment available at scale
Behavioral guidance built around a specific employee has always required knowing the manager and the employee well. HR has historically been the most scalable source capable of offering something so tailored, which meant that kind of preparation was only available to the managers HR had the bandwidth to reach directly.
A general-purpose AI tool can still only offer a manager a generic script pulled from broad patterns, without any sense of how a particular employee is likely to respond. But AI paired with validated behavioral insight works differently by drawing on data that contextualizes someone’s behavior, preferences and communication style. These details provide managers with guidance shaped around the individual they’re about to meet with, so they can walk into any conversation feeling more confident and prepared.
Confidence changes the outcome of a hard conversation as much as preparation does. Managers who know how someone is likely to respond can stay steady when the conversation gets uncomfortable, instead of losing their footing.
Here’s how AI with embedded behavioral context can improve the same common breakdowns mentioned earlier:
- Flagging an employee’s tendency toward defensiveness in advance lets a manager walk in ready for that reaction, avoiding the need to improvise after it happens.
- Surfacing how much context a specific employee needs before receiving criticism means a manager can have a more tailored approach and build that in from the start of the conversation.
- Showing how direct feedback tends to land with a specific person gives a manager language likely to be read as constructive while still delivering harder truths.
So how do organizations and their HR teams build that kind of behavioral infrastructure?
What HR should build first to scale behavioral support
Most organizations already have behavioral data sitting in assessment results, performance reviews and manager notes, but that data rarely gets translated into something a manager can use before a hard conversation.
HR’s first job is turning what’s already known about how someone communicates, processes feedback or responds under pressure into a clear framework for specific conversation types. That might mean a playbook for delivering critical feedback that flags when an employee needs processing time before responding, or one for managing team conflict that notes whether someone receives feedback better with context upfront or the issue stated directly.
Once those frameworks exist, AI built with more human context, grounded in validated behavioral assessment data, can carry them to every manager. That’s different from general-purpose AI, which has no read on the person a manager is about to sit across from. Behavioral intelligence gives HR’s frameworks somewhere to live, something a manager pulls up in the moment, without HR needing to be part of the conversation.
Starting with one conversation type, proving the guidance holds up, then expanding gives HR a way to scale behavioral support without waiting for a companywide rollout.
Leaders feeling confident in HR’s coaching and managers feeling supported are two different things, and more meetings won’t close the space between them. Closing it takes AI built with more human context behind it, something every manager can reach on their own in the moment they need it most.
The full picture of where leaders and managers disagree, and what managers say would actually help, is in the report. Get the 2026 AI and People Leadership Report.