The myth of the "technical talent shortage"

When a small business is asked why it has not made progress with AI, the most common answer is some version of “we don’t have anyone who understands this stuff”. It is a comfortable answer because it sounds like a technical problem with a technical solution: hire, train or wait for the market to settle. The trouble is that this is almost never the real cause.

What the data says, and what it leaves out

There is a real skills gap. According to Deloitte’s 2026 State of AI in the Enterprise report, a lack of employee skills is the most frequently cited barrier to integrating AI into existing workflows. Other analyses from the same year show that job openings related to AI are growing much faster than the pool of available professionals who can fill them.

But when this figure is repeated without context, an important nuance gets lost: most companies that cite this barrier do not need a team of machine learning engineers. They need someone who can translate what AI can do into a specific business process. That is an intermediary problem, not a shortage of PhDs in artificial intelligence. The distinction changes what makes sense for a small business to do.

Why “lack of talent” is such a comfortable excuse

A 2026 Harvard Business Review analysis of why AI projects stall makes a revealing point: many initiatives fail because employees’ anxiety about their relevance, identity and job security leads to superficial use without real commitment. It is not because the organisation lacks technical ability. When leaders treat adoption as a matter of psychology and context, not just a technical rollout, they are much more likely to turn experimentation into lasting impact.

Put another way, hiring someone with technical skills does not solve a problem that is fundamentally about incentives and trust. A Workera analysis of why AI adoption will remain uneven in 2026 says it directly: when people are rewarded for visible effort instead of outcomes, they will not automate tasks, whether or not they have the technical training. The talent shortage becomes a perfect excuse because it shifts the uncomfortable conversation (which process needs redesigning, who takes responsibility, which incentives need to change) to an easier one: “we need to hire someone”.

The small business case

At a large company, a shortage of technical talent is a real competition for scarce skills. At a business with ten or twenty employees, that is rarely the issue. The founder or operations lead does not need to become a data engineer or hire one. They need an hour of honest assessment to decide which process to automate first and why, plus an outside specialist to handle that technical part without requiring the company to maintain an in-house team all year.

When the conversation is reduced to “we don’t have technical talent”, it avoids the question a small business can answer without hiring anyone: which part of the current process shields someone from responsibility, and does that protection still make sense? A machine learning engineer cannot answer that. The person who knows the business can.

What changes the outcome

The 2026 data shows a consistent pattern across several sectors: organisations with formal AI training programmes report substantially higher proficiency and user satisfaction than those without them. But training is not the same as hiring, and that is where the costliest confusion lies. Training someone who already knows the process is often more valuable than hiring someone who does not.

A shortage of technical talent is a real obstacle in specific fields, especially model research and development. For the vast majority of small businesses looking to automate an administrative, sales or customer service process, it is not the main barrier. It is simply the barrier that is easiest to say out loud.