Mandatory AI adoption without adequate training isn’t a productivity strategy. It’s a trust problem.
AI adoption is increasing, but employee confidence in AI is lagging. Workers are being asked to use new tools, judge unfamiliar outputs, and adapt as their roles change. Without enough training, and without a feeling of psychological safety, that pressure can create anxiety instead of momentum.
That’s the AI trust gap: employees may use the tools, but they don’t always trust them, the process of adopting them, or their own ability to use AI well.
Workers are using AI more, but many don’t feel more confident using it.
EY research found that 71% of employees are concerned about AI. ManpowerGroup’s 2026 Global Talent Barometer found that regular AI usage had risen 13 percentage points, to 45% of workers, while confidence in using technology fell 18%, in the previous year. Pew Research Center found 52% of workers worried about AI’s future workplace impact, and 33% feeling overwhelmed.
Those numbers point to the same problem: more tools, less confidence. Employees who don’t trust AI, or their own ability to evaluate AI output, are less likely to use the tools well. They may use AI only for lower-risk tasks, or rely on AI outputs without knowing when to trust, revise, or reject them.
FOBO, or fear of becoming obsolete, is the anxiety that skills are losing value faster than employees can adapt.
This is different from fear of near-term job loss. FOBO is the sense that the work is changing while people are still expected to perform at the same level. Employees may not expect to be replaced tomorrow, but many still worry that their judgment, experience, or expertise is gradually becoming less valuable.
That creates a workforce that looks compliant in a utilization dashboard but still struggles to produce better results.
Many organizations roll out AI tools, set adoption goals, and expect employees to adjust quickly. Training, if it happens at all, is often a one-time event.
Research published in Humanities and Social Sciences Communications found that organizational AI adoption was associated with employee depression, with psychological safety playing a mediating role. When employees are expected to use AI without enough support, or the ability to admit uncertainty, adoption can create real strain that shows up as anxiety, disengagement, and declining job satisfaction. The organizations pushing hardest for AI adoption may also be creating the conditions that make adoption fail.
In our work with clients, the strongest returns for AI adoption and use come from organizations where employees understand the tools and feel equipped to make good judgments about how to use them.
When employees don’t trust AI or their own ability to use it, several problems show up that utilization dashboards may not catch:
If leaders only measure usage, they may see adoption rising and assume the program is working, even when proficiency—effectiveness in using the new tools—is low.
There are several rules you can follow to help both the adoption of AI in your workplace and the development of proficiency in using the technology.
Utilization tells you who’s using AI. Proficiency tells you whether they’re getting better at it.
Without proficiency data, training programs have no feedback loop. Managers can’t see who needs support, and leaders can’t tell whether AI adoption is improving output quality or just increasing tool activity.
Employees need to be able to say, “I don’t know how to use this well,” without it turning into a performance issue. Leaders can support that by modeling the sharing of uncertainty, normalizing questions, and making space for teams to share what’s working and what isn’t.
AI tools change too quickly for one-time training to work. Effective upskilling treats AI literacy as an ongoing practice, not a box to check at onboarding.
The goal is to help people build enough confidence and judgment to use AI well in the context of their actual work.
Larridin is built for AI measurement and optimization across the enterprise, including workforce readiness.
Larridin AI Fluency helps CHROs and HR leaders identify which teams are thriving, which are struggling, and where training investment will have the most impact.
Combined with Larridin AI Adoption and Larridin AI Impact, leaders can connect workforce readiness to AI usage, proficiency, and the creation of measurable value.
Book a discovery call to see how Larridin measures AI readiness across your workforce.
The psychological impact of AI on employees can include anxiety about job security, lower confidence when skills feel inadequate, fatigue from constant learning pressure, and reduced psychological safety when adoption is mandatory without support.
AI anxiety is the stress employees feel around AI adoption. It can include fear of displacement, uncertainty about how to use tools well, and worry that existing skills are becoming obsolete.
Build workforce trust in AI by communicating clearly about how AI will be used, providing ongoing training, and creating psychological safety so employees can admit gaps without fear. Measuring proficiency as well as utilization gives leaders the data to support people where they need it.
Utilization measures whether employees use AI and how often. Proficiency measures how effectively they use it and whether results are improving. High utilization paired with low proficiency is a warning sign that enduring value is not being created.
Want to know where your workforce stands on AI readiness?
Book a discovery call and get a clear picture of proficiency across your teams.