Navigating the AI skills shift – for everyone | The CEO Magazine

AI is changing what every role in the tech sector looks like – the skills needed, the pathways into leadership, and the way we develop the next generation. Getting this transition right is also one of our best opportunities to build a more inclusive industry. But only if we act with intention, because AI is also changing what it means to work in tech.

Those who help organizations adopt and implement technology now need to understand how and whether AI can genuinely help and how to put the right technology in place in the right way. Technology teams across the sector are working through how AI may affect their roles in the short, medium and longer-term.

The emerging view is that skills such as communication and stakeholder management will become increasingly valuable and that understanding a technical product end-to-end at a high level will matter as much as – if not more than – technical delivery expertise.

Those in technical roles will need to use these technologies to accelerate what they do, but they will also need to evolve into architects and quality assurers of AI outputs. And those focused on organizational change need to think about how to help businesses navigate these shifts, upskill their people and re-architect their operating models.

The boundaries between roles are blurring, and the need to hold a more holistic view of technology, ethics and business outcomes is now essential across the board.

New roles, new structures

There is a growing need for roles with deep skills on the ethics side, as well as in more technical AI and large language model architecture specialisms – and strong AI governance roles to ensure that increasing use of AI, both internally and in products and services delivered for clients, is underpinned by rigorous governance and compliance.

Junior roles will be redefined, focused less on doing the work themselves, and more on being skilled users and assurers of AI outputs. We need to place much more emphasis on becoming an expert in enabling AI change and adoption, or in AI ethics, data and model architecture.

These specific skills are not yet built into promotion pathways. We need to ensure junior staff are still able to develop the expertise needed to confidently assure AI outputs – without the experience of building technology solutions themselves in the way that their more experienced colleagues were able to.

This is a real challenge. We cannot simply replace junior roles with AI tools and expect a healthy pipeline of skilled mid and senior technologists to emerge. It may well require a deliberate business decision to invest time in training junior employees on tasks that AI could execute more efficiently, precisely to preserve the capability pipeline into more senior roles.

We also need to prepare for the reality that earlier-career applicants may arrive with more knowledge of AI tools than some of their more experienced colleagues.

Who enters, stays in and progresses

It is also essential that we encourage women and underrepresented groups at every stage of their educational and career journeys. We still see a divide at school: girls and boys perform similarly up to age 16, but after that, girls are much less likely to choose subjects such as physics, computing and engineering.

There appears to be a confidence gap – girls are less likely to believe they are good at subjects in the fields of science, technology, engineering and mathematics, even when their performance is equal. This is cited as being due less to a lack of individual confidence and more to structural issues. For women and other underrepresented groups, there is a lack of role models, a lack of expectation in schools and stereotypes persisting in the system.

Mentoring programs, communities, safe spaces to talk and authentic senior support at the point of entry are all essential in boosting diversity and inclusion in the workplace. Progression is key to retention, and for that, there needs to be a real focus on sponsorship and allyship, not just mentoring.

Many people leave their profession for caring duties or illness. Caring duties still largely fall to women, who then struggle to return. Eighty-seven percent of women leave the profession in the first 10 years of their careers, and caring duties are significant in this drop off.

Policies that support transitions in and out of work need to be strengthened. The recent focus on menopause policy is a good example of what is possible: it has opened a conversation that is enabling more women to remain in the workplace.

The evidence underpinning this picture is drawn from widely cited sources from the United Kingdom. There, women make up around 25 percent of the tech workforce overall, falling to 15–20 percent in senior roles. High attrition in the first five years is documented across multiple studies.

The picture for women of color is even starker – lower representation at entry and a steeper drop-off at every subsequent stage. Men, by contrast, enter at around 70–80 percent and maintain that advantage through to senior levels.

We also need to be able to recruit career-changers later in their working lives and ensure that entry-level initiatives are not geared solely towards school and university leavers. If we are to attract women and people from underrepresented groups into tech, we must be able to identify those with potential but without a traditional tech background.

That means being willing to take a risk on people who want to transition into tech without yet having the technical experience to demonstrate their capability – and investing in their training and development accordingly.

What actually works

Helpful interventions fall broadly into three areas.

First, inclusive policies: flexible working, parental leave, menopause policy, accessible recruitment and pay transparency.

Second, community: mentoring programs for guidance and support; sponsorship programs that actively uplift people; allyship; training; and both internal and external networks. Building an inclusive culture and creating spaces where diverse voices are heard and acted upon creates an environment where people feel engaged and are more likely to stay.

Third, process: hiring from the widest possible cohort, building clear training and promotional pathways, and ensuring that the forums shaping AI strategy have representation across levels and diverse groups.

Sponsorship is more effective than mentoring in allowing people to rise up in their organizations. We need active rather than passive mechanisms, and what gets measured gets done. Many past interventions were largely one-off and didn’t connect to an end-to-end process.

Targets and performance measures work. When we had a 30 percent women-on-boards target, we achieved it. Now, without sustained pressure, that progress is slipping. Transparency in policy and pay significantly influences how people experience their employment.

The gender pay gap figures are consistently better in the public sector, where transparency is mandated. That is not a coincidence.

What the data makes clear is that the biggest single drop-off point in the United Kingdom is not mid-career or at the point of promotion – it is at entry into tech roles. The second major loss comes in the first five years.

Women of color face compounded disadvantage at every stage of that journey. And the under-representation we see at senior levels is not simply a hiring problem – it is the cumulative result of attrition that begins right at the start.

A systemic challenge needs a systemic response – joining up changes in the school curriculum, the way tech is taught and the links between employers and schools and colleges that help build pathways leading to real jobs.

Policies at work need to be more inclusive and enable genuine flexibility. And sponsorship needs to produce more visible role models at the top of organizations – people who demonstrate that the art of the possible is real, and that it is possible for everyone.

Opinions expressed by The CEO Magazine contributors are their own.