As the next wave of artificial intelligence moves deeper into core business operations, organizational transformation requires expertise beyond traditional engineering. Navigating this technological wave demands continuous oversight across workflows, governance structures, and human capabilities to ensure long-term resilience. Discover how enterprise leaders can ride this transformational wave to build a future ready, inclusive ecosystem.
When a new technological wave hits the enterprise, organizational capabilities must adapt to the wave. To ride this wave successfully, leaders must manage the technical wave, the operational wave, the workforce wave, the governance wave, the adoption wave, and the strategic wave. If companies fail to prepare for this wave, the innovation wave stalls, the productivity wave declines, the skill wave recedes, risk management succumbs to the wave, decision quality collapses under the wave, and competitiveness falls behind the wave. When organizations anticipate each wave, align with the wave, and steer through the wave, enterprise transformation thrives across every level.
A recent post from Heather Tulk caught my attention. Responding to LinkedIn research showing that women accounted for just 26% of U.S. AI hires in 2025, Heather wrote: “There is no shortage of well qualified and capable women and yet we are repeating the sins of the past in tech.” The numbers give good reason for her concern. LinkedIn’s August 2026 research found that women represented just 26% of U.S. AI hires in 2025, compared with 50% of hires into non AI occupations.
Representation was even lower in some senior and highly paid positions, women accounted for 20% of Head of AI roles, while across the 27 countries examined in LinkedIn’s leadership analysis, women held only 13% of C-suite AI leadership positions at AI companies. LinkedIn Pressroom At the same time, the opportunity is expanding quickly. U.S. AI job postings have roughly doubled since 2023, and the typical AI job posting analysed by LinkedIn listed approximately $177,000 in compensation, compared with approximately $80,000 for a non AI role. LinkedIn Pressroom There is clearly a representation problem. But Heather’s comment made me wonder whether another problem sits underneath those numbers. Are we defining who belongs in AI too narrowly?
Look at many of the roles around which the AI labour market has formed. AI Engineer. Machine Learning Engineer. Member of Technical Staff. Director of AI. Head of AI. These roles matter enormously. We need people who can build models, infrastructure, applications and the technical systems that make AI possible. But something interesting is happening as AI moves beyond experimentation and into the operating environment of businesses. LinkedIn reports that AI Engineer has overtaken Machine Learning Engineer as the most common AI occupation on its platform. But Forward Deployed Engineer, a role centred on helping organisations implement AI solutions, is now the third most common AI occupation in U.S. job postings.
LinkedIn Pressroom More recent LinkedIn data makes the movement even clearer, demand for forward deployed experts increased 226% over the past year, while AI forensic analyst roles increased 169% and AI safety roles 91%. LinkedIn Pressroom Those are different signals of the same broader development. AI is moving out of the technology environment and into the enterprise. And once that happens, the problem changes.
An organisation can have excellent AI technology and still struggle to make AI work successfully across the business. Someone has to redesign workflows. Someone has to understand which processes should change and which should not. People need to learn new ways of working. Managers need to understand where responsibility sits. Governance has to evolve. Decision rights may need reconsideration. Existing expertise may be used differently. New capabilities need to be developed. Employees need to understand when to rely upon AI and when to challenge it. The organisation also has to determine whether the productivity gains promised by AI are actually becoming business value. These are not peripheral issues surrounding AI adoption.
They are increasingly the work of AI adoption. LinkedIn’s own research into executive leadership supports this broader picture. In June, it reported that 82% of surveyed C-suite leaders said AI was already creating roles within their organisations that did not exist a few years earlier, while 42% identified optimising workflows in which AI supports employees as critical. LinkedIn Pressroom Earlier this year, LinkedIn described an emerging new collar workforce combining technical skills with distinctly human capabilities, and reported more than 1.3 million new AI enabled jobs globally over the previous two years.
This is where I think the gender discussion becomes particularly interesting. There are already highly experienced women working across change management, organisational development, transformation, program and project leadership, governance, risk, human resources, learning and development, operations, customer experience and professional services. I am not suggesting women are inherently better at these things. Nor should women be directed toward the human side of AI while technical leadership remains predominantly male. Women should have every opportunity to become engineers, AI researchers, technical founders, Heads of AI and technology executives. But we may also be overlooking something much larger.
Many women already possess capabilities that enterprises are going to need as AI implementation moves deeper into the organisation. They may simply not think of those capabilities as AI expertise. And employers may not yet be looking for them in that way. A senior change leader who understands how people respond when workflows are redesigned may have considerable value in an AI transformation. A governance professional who understands accountability, escalation and decision rights may have considerable value in an AI enabled operating model. An organisational development specialist who understands how expertise is formed and maintained may become increasingly important when AI begins performing work through which people previously developed that expertise.
A transformation executive who understands how to move an organisation from technology deployment to sustained operational change may be as important to the success of an enterprise AI program as some of the technology itself. None of these people needs to pretend to be a machine learning engineer. The opportunity is to recognise that enterprise AI needs more kinds of expertise.
I know something about this personally. I didn’t enter AI through computer science or machine learning engineering. My work began with a different question: what happens to humans and organisations as increasingly capable AI becomes part of how they think, work and make decisions? That question eventually led me to become a Human AI Architect and to build Gaia Nexus’s proprietary behavioural governance architecture. Our work looks at what happens after AI becomes part of everyday work whether people begin relying on it differently, checking it less, challenging it differently, exercising their own judgement less often, or gradually changing how decisions are made.
At Gaia Nexus, I have developed a suite of Behavioural Protocols, also described as AI Agent Instructions, to help organisations identify and understand these changes as AI becomes more embedded in their operations. Suite of Behavioural Protocols I can also turn identified changes in those conditions into structured Detection Events that external systems can consume without receiving our protected behavioural protocols or derivation logic. I did not find an existing profession and decide to enter it. The work emerged because the problem emerged. And I suspect my experience will become increasingly common.
Consider what happens after an AI implementation is declared successful. The model performs well. The workflow becomes faster. People use it. Costs decline. Productivity rises. From a conventional technology perspective, this can look like success. But what is happening to the people inside that workflow? Perhaps employees verify AI recommendations less frequently because experience has taught them that the system is usually right. Perhaps people who once developed expertise by performing hundreds of lower stakes decisions are now supervising AI instead. Perhaps an experienced professional still holds formal approval authority but increasingly receives a decision that has already been substantially shaped upstream by AI.
Perhaps productivity improves while the organisation becomes more dependent on expertise that fewer people are continuing to develop. None of these possibilities means the AI deployment has failed. They mean the organisation has changed. And understanding that change is part of governing AI successfully. This is why I believe enterprise AI will increasingly require people who can work across technology, operations and human systems rather than exclusively inside one of them.
There is another implication here. Traditional change management often helps an organisation move from one operating state to another, introduce the technology, redesign the process, train people, manage adoption and embed the new way of working. AI complicates that model. The technology continues changing. People continue adapting to it. The relationship between the two can continue changing long after implementation. A person who was appropriately cautious during the first month may behave very differently after twelve months of consistently accurate AI recommendations. A team may become faster and more capable with AI while simultaneously exercising some independent capabilities less frequently.
A control that was meaningful when a system was introduced may still exist a year later while the human behaviour surrounding it has changed considerably. So AI transformation cannot always be treated as a transition with a clean end point. The organisation may need to understand change continuously. That potentially creates an important evolution for change management, transformation and organisational development professionals themselves. Their role may not end when adoption succeeds. It may increasingly include understanding what successful adoption subsequently changes.
This becomes particularly important when we think about the next generation of workers. If AI removes repetitive or lower stakes work, that can be an enormous productivity gain. But organisations should also ask what else that work was doing. Was it how junior employees learned? Was it where people developed pattern recognition? Was it where professional judgement was formed? Was it where future leaders accumulated the experience they would later need for consequential decisions?
If AI removes the work, where will the human capability that work previously created come from? That is not an argument against automation. It is an argument for understanding the organisational consequences of automation well enough to redesign capability development alongside it. And again, solving that problem will require expertise well beyond AI engineering.
LinkedIn’s research should concern us because women are clearly under represented in many of the economic opportunities being created by AI. We should address that directly. But I don’t think the answer is simply to encourage more women to become engineers. It is also to recognise that the definition of an AI career is expanding. The enterprise AI workforce of the next decade may include engineers, data scientists and technical architects alongside transformation leaders, governance specialists, organisational designers, educators, operations professionals, risk specialists, behavioural experts and people responsible for maintaining human capability in increasingly AI enabled organisations.
Some of those professions may not even have names yet. And some of the people most qualified to shape them may not currently consider themselves part of the AI industry. That is the opportunity I think we should be talking about. Not lowering the threshold for what constitutes AI expertise. Expanding our understanding of the expertise that successful AI actually requires. Heather is right: there is no shortage of capable women.
Perhaps the next question is whether the AI industry and the enterprises investing billions in AI transformation, are looking broadly enough to recognise the capabilities already sitting in front of them. Because the next phase of enterprise AI won’t only be about who can build the technology. It will be about who can help an organisation change around it without losing the human capabilities it still needs. And that could open the door to a much broader generation of women shaping what AI becomes.
