How Do We Handle Employees Resistant to AI Training?

Every financial services firm rolling out AI literacy training eventually meets the same problem.

Some staff complete the modules without complaint. Others quietly avoid them, raise objections in team meetings, or comply only on paper while remaining disengaged.

Given regulatory expectations and government workforce initiatives such as the Skills Compact, AI training is no longer optional. But mandating attendance does not resolve the reasons staff resist it in the first place.

Understanding why resistance happens, and responding to the specific cause rather than the symptom, is what separates a training rollout that genuinely builds capability from one that simply ticks a compliance box.

Why Employees Resist AI Training

Resistance to AI training rarely stems from a dislike of learning itself.

In financial services specifically, common underlying causes include:

• Fear of redundancy. Staff in operations, settlements or customer-facing roles may reasonably suspect that training on AI tools is a precursor to their role being automated.
• Scepticism about relevance. Experienced professionals who have seen previous technology promises fail to materialise may doubt that this initiative is different.
• Change fatigue. Staff who have been through multiple system migrations, restructures or compliance overhauls may resist simply because another change programme has arrived.
• Lack of confidence with technology. Some employees, particularly those without recent exposure to new software, may worry about appearing incompetent in front of colleagues or managers.
• Lack of psychological safety. Where firms have not been transparent about the purpose of AI adoption, staff may assume the worst and disengage as a form of self-protection.

These causes often overlap, but they call for different responses. Treating all resistance as a single discipline problem, to be solved by mandating completion, misses the underlying issue and can damage trust further.

Traditional Approaches to Overcoming Training Resistance

Long before AI entered the picture, organisations developed established methods for addressing resistance to workplace training and change programmes. These remain the foundation of any response.

Manager sponsorship. Staff take training seriously when their direct manager visibly participates and endorses it, rather than simply forwarding a link to an e-learning module.

Transparent communication. Explaining clearly what the training is for, what it will and will not lead to, and how it fits into the firm's wider obligations reduces speculation and anxiety.

Staged rollout. Introducing training in phases, starting with teams or individuals more likely to engage positively, allows early advocates to influence sceptical colleagues.

Addressing concerns directly. Rather than deflecting questions about job security or workload, giving a direct and honest answer, even an imperfect one, builds more trust than avoidance.

Peer advocacy. Colleagues who have already engaged with the training and can speak to its practical value are usually more persuasive than management messaging alone.

These principles apply to any workplace training initiative. AI training is not exempt from them, and no AI-specific tool replaces the need for this groundwork.

Where AI-Supported Approaches Help

While change management fundamentals remain essential, AI-specific tools and methods can address resistance in ways generic approaches often cannot.

Hands-on, low-stakes exposure. Allowing staff to use a sandboxed AI tool on familiar, low-risk tasks, rather than reading about AI in the abstract, often does more to reduce scepticism than any amount of instruction. Seeing directly what a tool does, and does not do, replaces speculation with evidence.

Adaptive learning paths. Rather than delivering the same generic module to every employee, adaptive systems can adjust pacing and content based on an individual's existing familiarity and role, reducing the frustration of overly basic or overly technical material.

Engagement and sentiment monitoring. Structured feedback tools and engagement analytics can help surface patterns, such as a particular team consistently disengaging, that might otherwise go unnoticed until completion deadlines are missed.

These tools support the diagnosis and delivery of training. They do not replace the judgement required to interpret why a particular team or individual is resistant, or decide how to respond. That judgement remains with managers and HR business partners.

Practical Considerations When Running the Rollout

A few practical steps make a measurable difference to how resistance plays out during a rollout.

• Segment training by role. A settlement clerk, a relationship manager and a compliance analyst have different reasons to care about AI, and different concerns about it. A single organisation-wide module tends to feel irrelevant to at least some of these groups.
• Make managers visible participants, not just enforcers. Managers who complete the training publicly and discuss it openly signal that it is taken seriously, not simply imposed from above.
• Avoid over-promising. Training should not be framed as guaranteeing job security or eliminating all uncertainty about AI's impact on roles. Overstating reassurance that later proves untrue damages trust more than acknowledging genuine uncertainty upfront.
• Monitor whether resistance eases over time. Completion rates alone do not indicate genuine engagement. Following up with informal conversations or short feedback surveys helps establish whether underlying concerns are actually being addressed.

AI training rollouts succeed or fail based on how well they are managed as a change initiative, not on the sophistication of the training content itself.

Example

A mid-sized London bond trading desk is rolling out mandatory AI literacy training ahead of new internal governance requirements.

Several experienced settlement staff quietly avoid completing the online modules, and one raises concerns in a team meeting that the training implies their roles will be automated.

The Head of Operations pauses the generic rollout and asks the L&D Business Partner to run a short, role-specific session using a sandboxed AI reconciliation tool so staff can see directly what the tool does and does not do.

The session addresses the redundancy concern openly, and completion rates for the wider training programme improve once staff see the tool as something that removes repetitive checking rather than replaces their judgement.

FAQs

• Is resistance to AI training usually about job security?

  Job security is a common driver but not the only one. Scepticism about the training's relevance, fatigue from previous change programmes, and general lack of confidence with new technology are all frequent causes. Each requires a different response, so it is worth understanding the specific concern before responding rather than assuming redundancy fears are always the root cause.

• Should we make AI training mandatory if staff are resistant?

  Mandating training may be necessary for regulatory or workforce capability reasons, and firms are increasingly expected to do so. However, requiring attendance does not resolve the underlying resistance. Firms should pair any mandate with genuine engagement efforts, such as transparent communication and hands-on exposure to the tools involved, rather than relying on the mandate alone.

• Can AI tools themselves help identify where resistance is coming from?

  Engagement analytics and structured feedback tools can help surface patterns, such as particular teams or roles showing consistently lower completion or engagement. However, interpreting why that pattern exists and deciding how to respond still requires manager judgement; the tools identify where to look, not what to conclude.