AI in engineering will need the human touch.
The conversation around AI should not just be about the technology itself, but how it’s applied and who’s applying it says Simon Farnfield, event director for Advanced Engineering UK.
Much like the early conversations around Industry 4.0, there’s still a sense that many companies aren’t entirely sure what AI is or how it can benefit them. That’s concerning, because in engineering and manufacturing, AI is no longer a distant prospect.
91 per cent of companies are planning to increase their investment in AI over the next two years. Although investment is ramping up, there’s a growing concern that the skilled people needed to manage and scale these systems aren’t entering the workforce fast enough.
Part of the problem is a lingering misconception that AI is something you can simply buy off the shelf. Reality is, AI is a tool that only delivers value when it’s supported by the right strategy, quality data and skilled talent.

Why human skills matter
It’s true that AI will replace some repetitive tasks, but that doesn’t necessarily mean jobs will disappear. Instead, it’s changing what those jobs look like. As lower-value tasks are automated, people will be freed up to focus on higher-level decision-making and creative thinking.
That’s where the opportunity lies, but only if the workforce is equipped to make the most of it. Before we get into the technical skills required, soft skills like communication and leadership are often overlooked in this discussion, but they’re more important than ever. While AI can crunch numbers or generate solutions, it can’t negotiate, lead a team or build trust with stakeholders.
These are still distinctly human skills, and they’re critical, especially at senior levels. Leaders in engineering don’t need to become coders, but they do need to understand how AI works, what its limitations are and how to ask the right questions when making strategic decisions.

Agentic AI
This is where blind adoption becomes a huge part of the equation. AI systems are built on data, and that data is often biased — sometimes subtly, other times significantly. Without that layer of human oversight, there's a risk that AI could lead us down the wrong path — and faster.
This is important as we move into the next phase of AI, called agentic AI, which McKinsey names as the “next frontier.” These aren’t just systems that follow instructions, they’re designed to act independently, make decisions in real time and respond to complex environments without waiting for human input.
But with that comes greater responsibility and the need for careful human guidance. Trust in AI must be earned, which depends on having the right people involved at every level — not just to monitor what the system is doing, but to dictate what it's doing in the first place.
Agentic AI has potential, but like any tool, its value is only realised when it's used thoughtfully and with the right checks and balances in place.

The mismatch
To meet this challenge, industries will need people who understand data. Programming and machine learning will be crucial, not just in developing AI tools but also in maintaining and improving them.
There’s often a lag between what companies need and what schools, colleges and universities are teaching. While apprenticeships are beginning to address the shortfall, there’s still a lack of clarity and direction across the wider education system.
Coding and data science need to be embedded much earlier in the curriculum. That’s not just at university level, but in secondary schools, T Levels and vocational training programmes too. These aren’t just nice-to-have skills, they’re becoming on par with literacy and numeracy in a world increasingly driven by digital systems.
Too many young people also think of engineering as dirty, manual work, yet AI offers a chance to reframe that perception. Today’s engineering roles involve an incredible amount of creative work using robotics, sensors, software and data science to design and operate today’s innovative technologies.

Get guidance
In the meantime, companies must take responsibility for building their own future workforce. That starts with understanding which data matters most to their operations.
Small-scale pilot projects are an effective way to begin, trialling AI on a specific process or production line to learn what works and where internal gaps exist. These early experiments not only help know how AI can help, but also identify the specific skills needed to scale AI successfully.
Support from organisations, like Made Smarter, offer training, online courses and funding for pilot projects that help companies begin their AI journey in a manageable way. Likewise, Advanced Engineering UK’s new ‘Unsung Heroes in Manufacturing’ initiative, a part of #MINDTHEskillsGAP, provides easier access to the real-world voices to connect upcoming engineers with industry experts.
These efforts matter because AI’s true value comes from learning how it’s best applied, rather than being just about the technology itself. It’s a tool that only works if you understand the problem you’re trying to solve.
That’s why the companies that succeed in the next phase of AI adoption won’t be the ones with the smartest machines, it’ll be the ones that understand the data they value, with the right people guiding how it’s used.