A new employer survey finds that artificial intelligence itself is not what decides whether workers see the technology as a career boost or a threat. Whether their employer bothered to train them on it is doing that work instead, and by Businessolver’s count, nearly half of employees are being left to figure AI out alone.
The training gap is doing the real work
Businessolver’s 2026 State of Workplace Empathy AI Special Report, built on a survey of more than 1,300 U.S. employees and executives across six industries conducted with Edelman Data & Intelligence, found that 49% of employees received little or no employer-sponsored AI training. The gap in outcomes between the trained and the untrained is stark: 78% of adequately trained employees say AI has accelerated their careers, against 51% of those with limited training. Confidence follows the same pattern, with 80% of trained employees saying they feel able to keep up with AI at work compared with 57% of the undertrained.
Trained employees are up to 1.5 times more likely than untrained peers to report career growth, confidence, and optimism about AI. That is a sizable return for what the report frames as a relatively cheap intervention next to the AI platforms themselves.
CEOs are investing in AI for two different reasons
The report also surfaces a split at the top. Twenty-eight percent of CEOs cite headcount reduction as the primary driver of their AI investment, while 31% name upskilling their workforce as the primary goal. Those are two different bets on what AI is for, and they produce two different employee experiences downstream: one where AI is a tool workers are equipped to use, and one where it is a signal that their job is next.
Even so, 60% of C-suite respondents say AI adoption requires a “much greater need to demonstrate empathy,” and 61% of CEOs say they personally find it difficult to demonstrate empathy day to day, up 32 points year over year. Leaders are naming the requirement faster than they are meeting it.
Toxic cultures are winning on paper, for now
The report’s most uncomfortable finding is that dysfunction currently pays: cultures CEOs describe as toxic are delivering financial growth at roughly twice the rate of empathetic ones, even as 33% of CEOs (up 25 points year over year) and 40% of employees now describe their own culture as toxic. But the training data suggests that edge will not survive the AI transition intact. In empathetic cultures, 87% of employees received adequate AI training; in toxic cultures, only 33% did. Businessolver CEO Jon Shanahan called the toxic-culture growth numbers precursors of “a cultural reckoning,” not a stable advantage.
The gap is not evenly distributed
Training access breaks along familiar demographic lines. Women reported adequate AI training at 45%, against 56% of men. Gen X employees were the group most likely to say they were left to figure AI out alone, at 53%, and 41% reported career disruption from AI, the highest of any generation in the survey. Employees in hands-on, non-desk roles trailed their computer-based peers by 13 points on training access. Chief AI Officer Sony SungChu put the design principle plainly: “AI is most powerful when it amplifies human potential, but people need guidance.”
The mental-health data underlines the stakes of getting this wrong: 64% of employees who are anxious about AI reported mental-health strain tied to it, compared with 50% of employees who are optimistic about AI. Training is not just a productivity lever here; it is functioning as a wellbeing intervention.
What it means for the HR leader
The report reframes what “AI readiness” means for HR. It is not a model-selection question or a rollout timeline; it is a training-budget and manager-empathy question, and the data suggests HR functions are already being scored on it by employees. Two threads from HRTech’s own coverage point the same direction: leadership readiness, not technology readiness, has become the real bottleneck in AI adoption, and separately, employees are already adopting AI faster than employers can govern it. Put together with the Businessolver data, the pattern is consistent: employees are moving on AI whether or not HR has a plan, and the plan that is missing most often is training, not policy.
HR leaders should also treat the demographic gaps as a compliance and retention risk, not just an equity talking point. A 13-point training gap between desk and non-desk workers, or an 11-point gap between men and women, is the kind of pattern that shows up later in attrition data and, potentially, in disparate-impact scrutiny of who got AI-adjacent opportunities and who did not.
What to do now
Four moves the data supports. First, audit who has actually received AI training, broken out by gender, generation, and desk versus non-desk role, since the survey shows the gaps are structural, not random. Second, brief managers specifically on empathetic delivery of AI-related change, since CEOs are naming empathy as a requirement faster than they are building the skill in themselves or their leadership layer. Third, treat AI training completion as a wellbeing metric alongside a productivity one, given the mental-health gap between confident and anxious employees. Fourth, if AI investment is being justified internally as a headcount play, say so plainly to the workforce rather than wrapping it in upskilling language; the survey suggests employees can tell the difference, and the trust cost of a mismatched narrative compounds every quarter it goes uncorrected.
The technology rollout is the easy part. The training rollout, and the honesty about why the technology is being bought in the first place, is the one the data says most employers are still failing.
Source: Businessolver
