The Water Is Getting Warm. Your Best People Can Feel It.
Unease: AI is creating simultaneous excitement and career fear, making employee sentiment a critical organizational signal.
Tooling: Tool restrictions can drive shadow usage, frustration, and attrition when employees cannot access effective systems.
Listen: HR needs continuous, trusted feedback to identify tooling gaps, learning needs, and concentrated pressure before departures.
Equip: Role-specific AI education builds practical judgment, responsible use, and confidence beyond what compliance training can provide.
Accountability: Clear standards become fair only after employers teach workers how to verify AI outputs and handle failures.
Somewhere on your team, or in a team you support, is a person who is very good at her job and has been on a steady course of acceleration for ten years. Let’s call her Margo.
This year, for the first time, she is not sure what her job is becoming. On Tuesday, she used an AI tool to build a small automation that saved her group a day of work, and she felt like the sharpest person in the building.
By Thursday, she read that a competitor now has systems doing a third of what her department does, and she lay awake doom-scrolling through jobs on LinkedIn and doing arithmetic about her own relevance and career path. Both of those things happened to the same person in the same week.
Create a Free Account to Keep Reading—and Keep Leading Smarter
Unlock this piece and join a community of forward-thinking leaders discovering tools, playbooks, and insights for thriving in the age of AI.
Have an account? Log In
Finding Footing on Moving Ground
That is the uneven texture of professional life across almost every discipline right now, and a conversation about employees and AI that starts anywhere else misses the chance to understand what matters most to your organizational strategy. There is no field of focus or study that is not shifted by the advent of the broad range of tools and predictive systems we now call AI, and there is no function or team in your organization that is immune to the incendiary mix of pressures and opportunities that these technologies bring with them.
At every level within your organization, professionals are trying to orient while the ground moves. Those trying to keep pace seem like cowboys who need policing, and those trying to stay the course until the organization plays the changes look like they need to be brought up to speed, but every one of them is playing by ear.
If you lead an HR function, that Tuesday-to-Thursday swing is an operating signal, and it is telling you something specific about tool access, workforce enablement, and attrition that you could still help prevent.
Consider what Margo is working to navigate. The foundations of her discipline are shifting, and the skills that made her the person everyone asked for over the last ten years are proving to be replicable through tooling.
She trained for and built a career that your industry may not rely on as heavily in the years to come, and she is both excited about how the tools can help her teams and wary of how they may encroach on her function.
Colleagues doing Margo’s job at other companies talk openly about tools she is not allowed to touch, and their work is visibly faster for it. From her own leadership she gets one of two things: pressure without direction, or a silence so complete it functions as a message.
Underneath all of it sits the request that colors everything else: contribute what you know, your shortcuts, your judgment, your institutional awareness, to systems that will streamline the work for everyone.
Duncan, who came on board just a few months ago, is chewing the knuckle of his index finger in meetings and thinking of taking up smoking again as he tries to process that the company has only one authorized AI tool suite and his go-to tools are only going to be usable off to the side.
His portfolio is what got him in the door and it was built with tools that now sit outside it. He is starting to understand that he may have been hired for an unseen AI fluency that he will not be allowed to use.
Margo and Duncan are feeling the different ends of the same current. Her discipline is being absorbed into tools she may be asked to help train, he is locked out of tools he is not allowed to touch. Maybe you are never going to replace her, and maybe his tool list doubles by spring. If that is the case, say so, and find ways to show that you mean it, but do not imagine that it settles everything.
The uneasiness is in pockets across every function, and proves cumulative. No single assurance can resolve what no single event caused. The work of organizational culture is always called upon to account for and offset things that it cannot fully see, but with the pace of change AI has created, it needs more information to be effective.
What Uneasiness Produces
My father used to say that all bad decisions are made from fear. I don't think he was right.
Plenty of bad decisions are made from magical thinking or hubris, and a fair number from simple exhaustion. But he was pointing at something worth looking at, as fear often produces seemingly irrational decisions that feel rational because they are made from within a distorted environment.
Fear of not doing their best work will cause some people to reach for the tools that work best for them, sanctioned or not, because their professional standards will not let them fall short of a standard they know is possible.
Some dig in and avoid all such tools. They entrench in the part of themselves the organization has always valued, treating their expertise as the last defensible ground. (A trench is a complicated thing within a workforce, it shelters you at first, but it gives you a very limited vantage point and means of interaction with what is going on outside of it.)
Still others will run the numbers on the whole situation and conclude that whatever they need to be learning or doing next, they will not be learning it or doing it here. Those are the ones already returning recruiters' calls, and they are disproportionately going to be the ones you can least afford to lose, because appetite for the new and marketability travel together.
The banned tool, the entrenchment and avoidance of these new technologies, the resignation letter may each look irrational from the HR seat that is asked to enforce policy, but each one is a reasonable answer to the situation as it appears from inside it.
There is a vertigo specific to being held in the air by a seesaw whose other side carries all the weight, and here the weight is both power and information.
Leadership holds the AI strategy. The plans, the tool decisions, the AI literacy road map (or the assumption that people will upskill themselves), the productivity targets and the AI-derived revenue targets, and what the roles are meant to become.
Employees hold the career risk those decisions create. When the party with the least information carries the most personal exposure, unease is rational. Silence is what keeps the imbalance in place.
Where your organization controls most of the information, silence, the unexplained bans, literacy and upskilling opportunities, limits, or absence, and any automation being sought or established create the environment your people are reading and navigating.
The more they feel the water warming around them while being asked to pour institutional knowledge and personal insights into the tools, the closer they move to one of the exits, whether away from the technology or away from your organization.
The First Obligation (and your biggest opportunity) Is to Hear Them
You cannot answer to an uneasiness you cannot hear, and most organizations are not set up to hear it. Similarly, organizations cannot adjust their course and strategy to the problems they cannot see, and without a means to understand, on an ongoing basis, the pressures acting upon their employees with regard to AI, they will remain blind to the risks up ahead.
What you need to know
The tooling gap
What people believe they need to do their work well that they are not permitted to use, which is exactly what unsanctioned tool use has been trying to tell you all along. Also why the tools you chose for them are insufficient, or why the tools they lobbied for and have licenses for are no longer the best in class.
You need this information in real time, so that you can adjust messaging to staff to explain and reinforce your choices, or to make changes to vendors and subscriptions.
The education gap
What people are being asked to work with that nobody has taught them to use to their full advantage, which currently surfaces as anxiety and quiet errors. What they feel they need to learn next, as this is a continuum of learning and adult education, not a space in which a click-through compliance training can upskill people effectively.
The pressure gap.
The places you are pushing harder than people can absorb while their professional ground is moving beneath them, the places you are not moving fast enough, or where your AI strategy is not aligned to the way your business runs, all of which otherwise will surface as departures your exit interviews will politely misattribute.
Are You Listening?
Some people are trying desperately to get this information to you, but they aren’t being heard where they raise it, or they do not know where to raise it, and the instruments most organizations use cannot carry it.
The annual engagement survey arrives too rarely to catch anything in motion, and no one trusts its assertions around anonymity. The open door requires exactly the courage the atmosphere caused by AI is eroding daily.
Your security team and tooling dashboards can tell you which tools people touch, but surveillance primarily teaches people to hide better, and it does not carry answers as to why. Exit interviews, even when they are honest and thorough, arrive after the loss.
What these all have in common is that they leave silence undisturbed, and silence gets read by leadership as calm, despite all that lies under the surface.
Your organization needs to see that its teams are the cure for its blind spots, here, and it needs to want and solicit feedback on its AI use and strategy on an ongoing basis. It needs to cultivate a feedback culture with regard to these technologies.
Your people need a safe way to give you candid feedback and updates that provide you with aggregate insight. You need a way to hear from them and to know where pressure is concentrating, where tool restrictions are creating impasses to progress, where intentions to leave are clustering without knowing who it is coming from, because you do not need who. You need trust.
What a functioning feedback culture and mechanisms tied to AI give you is what I call “Institutional Sight”, or the capacity to see what is happening within and understand the pressures acting upon your organization while there is still time to make choices about what to do.
Without it, you are managing the version of your company that your leadership intended, even where the ground beneath its stakeholders has shifted it further and further away from its stated strategy.
The Second Obligation Is to Equip Them
Something close to new is being asked of employers now. Professional development used to mean deepening people within a stable discipline or better orienting them to their industry.
What this moment demands is role-specific adult learning across nearly every function at once, while the tools and the work change underneath it, and few employers have ever had to sustain anything like that.
The generation growing up with these systems will arrive fluent, but fluency is not judgment, and organizational standards and verification protocols will still have to be taught even to them. But until AI-native workers are a much larger share of your workforce, even the fluency is missing, and the whole orientation burden sits with the employer. There is nowhere else it can live.
Equipping people to understand the many tools that fall under the broad definition of AI, what they can do, and where they fail, is a moving target as the tools continue to evolve. People need to understand enough to use these tools to come up with realistic ideas and useful solutions in their own pockets of the organization, where most real adoption happens.
They need to understand the limitations well enough to know why and when the output cannot be trusted. And they need to know and understand that responsible use is a skill they require as a professional.
A one-hour compliance module can introduce the rules you have set. It cannot build the judgment to apply them, or the fluency that creates the innovation and efficiency you are asking for. Learning built around what these tools and their limits mean for each discipline and team, revisited as the tools change, can, and it does something harder at the same time: it answers the uneasiness and changes the atmosphere.
An organization visibly investing in making its people more capable has shown them it intends for them to stay. It has also built the capability that innovation, efficiency, and compliance depend on.
The Standard That Education Makes Fair
Equipping people is also what earns you the right to hold them accountable. Put the standard to your workforce plainly. If you handed your work to a brilliant twelve-year-old and delivered whatever came back without checking it, and it was wrong, you would be fired.
The job is yours and the work is your responsibility. Working with AI is no different than handing your work to a brilliant twelve-year-old who you think you have explained things to very clearly. While working with AI is authorized and encouraged, accountability and judgment are what you are relied upon for.
That standard is fair, but only in sequence. It assumes people know what checking means for these systems, where the failures hide, what confidently wrong output or improper use looks like in their specific team or field of work.
Enforce it without teaching any of that and you have built a trap and called it a policy. Create a real path to AI literacy and the same standard becomes a professional competence people can be proud of meeting.
No One Is Better Placed Than You
The reflex will be to route or defer all of this to the AI lead, IT, or legal, or a committee that meets quarterly, because they are handling AI and the attendant change management.
Human Resources professionals should resist that urge to stand back, because the feedback culture and channels yield a tooling needs map, an education demand signal, and an attrition forecast, which is HR information by nature. It is exactly what HR is best positioned to help the business explain to its teams.
Legal, IT, and Security will build the pieces of the machinery around safeguards, approvals processes, and policy language. Leadership will have to build the strategy around tooling limitations, efficiencies, AI literacy, and revenue targets.
What they cannot always see clearly is the need to make the feedback channel or the learning feel safe. Engagement and safety are functions of workforce trust, and workforce trust is an asset HR is uniquely positioned to see and shift. A chief human resources officer who waits to be handed a role in this will eventually be handed the fallout instead.
The uneasiness in your workforce is not going to dissipate and self-resolve in silence. Left alone, it converts into those shadow AI tools, trenches, and resignation letters. Answered, with a channel people trust and an education that feels like an investment in them. It converts into people building solutions in their corner of the company and telling you about them, people flagging what isn’t working as well as you think, and people contributing what they know because they know they are safe to be part of the chorus of voices shaping strategy.
Your people are deciding right now which of those organizations they work for, and they are using the information you have given them.
