Hayley McKelvey
Chief AI Officer, Deloitte UK
Businesses must not lose sight of strong leadership and the value of their human workforce as they strive to build AI into the fabric of their organisations, according to a leading expert in the field.
Direction from the C-suite — either from the CEO or a designated executive taking responsibility for the introduction, use and objectives of AI — is essential. According to Deloitte UK’s Chief AI Officer Hayley McKelvey, this defines AI’s deployment while reassuring the workforce that may be apprehensive about its adoption.
AI depends on leadership
“The AI opportunity is a leadership, rather than a technological challenge, but that is sometimes not fully grasped,” she says. “That’s why organisations need to have that job at an executive level.” McKelvey suggests that companies that put AI on the executive agenda make “demonstrably more progress.”
“AI is incredibly systemic and goes through the veins of an organisation. It needs somebody with permission and experience to look across the whole organisation and think about how to affect change coherently across a multitude of different places.” However, reality paints a different picture. A 2025 Deloitte survey of 1,854 executives showed that only 10% of companies have their CEO as the primary AI lead.1
Scaled impact
With the ability to connect trusted data, provide analytics and offer AI-enabled automation, AI implementation has become an imperative, with 96% of FTSE 100 companies referencing it in their annual reports, though the maturity level of those organisations differs dramatically.2
“That is often relative to size, the level of regulation and technological debt, as well as risk appetite and the ability to transform,” she adds. “When it comes to people, workers are definitely keen to experiment with AI and are using it quite prolifically, but most organisations are still on a journey of connecting that use, curiosity and experimentation and turning it into scaled impact across the business.”
McKelvey has recently stepped into the role of Chief AI Officer at Deloitte, having been with the company for 23 years and previously overseeing AI in its Tax and Legal business. Her appointment — the first at C-suite level in Deloitte — highlights the importance of high-level leadership in the AI arena.
Workforce survey
Deloitte closely monitors AI growth in the workplace, with its GenAI Workforce Survey of 25,000 UK workers highlighting how employees use the tool and how organisations are responding.3
Some 63% use generative AI (GenAI) for work, including drafting emails, searching for information and creating summaries. Of these, 46% are using free tools, 34% use external tools paid for by their employer and 17% use in-house tools. But some workers pay for their own GenAI tools for work, showing that some employees are embracing the technology quicker than organisations are making it available.
Clear AI vision empowers employees
Additionally, 65% report a lack of convincing leadership on how GenAI should be used in their organisation, prompting McKelvey to suggest companies must be clear on why they are using AI. “Unless you can explain why you are doing it, what it should look like and to what end, it is very difficult for organisations to create enterprise value out of the use of AI,” she adds.
That clarity of vision can empower employees and reduce fear of AI, particularly among those feeling stressed or anxious about their jobs. As companies strive to turn AI ambition into measurable business value, employees should be given the space to learn to use the tools effectively.
“You cannot shortcut this,” she says. “Learning to work with AI depends on the job you’re doing; one size does not fit everyone.” Consequently, leaders who understand what the business is trying to achieve, stratifying where they can use AI and for what end, breed confidence throughout a company.
Efficiency, innovation, resilience
“It is so important for organisations to know what they stand for when it comes to AI,” she adds. “AI represents a significant opportunity for many brilliant things to happen but also has a lot of scary prospects around it. I think it is possible to feel both optimism and trepidation at the same time.”
Organisations must think consciously about how they bring AI into the workplace, balancing AI and the human workforce. “We should not surrender human judgement to AI systems,” explains McKelvey. “Creativity, ethical judgement and decision-making remain fundamentally human strengths. AI can support these activities, but it should not replace our role in them.
“If you take the things that are uniquely human with those that are incredibly powerful about AI and push them together, that is where value lies and where organisations, and their people, will do really well.” As AI becomes more prevalent, McKelvey believes that human relationships become more important.
“It might feel paradoxical, but as businesses increasingly weave AI into their fabric, and people work with AI colleagues, the need for human relationships, proximity and empathy arguably increases rather than falls away,” she continues. “You are asking humans to become increasingly polymathic — to be great at interacting with AI but also retain what they are good at as a human and not just taking what AI says as a given.”
AI costs are rising
Cost remains a factor in deciding AI deployment. Deloitte say 85% of organisations increased their AI investment in 2025, but ROI remains at two to four years compared to the typical payback period of seven to 12 months expected for technology investments.1 Across industries, generative AI has become the fastest-growing line item in corporate technology budgets, already consuming up to half of IT spend in some firms.4
Companies must measure and manage the total cost of AI ownership. Deloitte helps clients to understand the cost and the rapidly developing economics of AI. AI cost is shaped by what you run and how you run it. Deloitte’s recently launched open model engineering practice specifically supports clients to scale agentic AI using a mix of open and proprietary models.
leaders who understand what the business is trying to achieve,
stratifying where they can use AI and for what end, breed confidence throughout a company
AI trustworthiness
With its immense power and the ability to enable businesses to harness data, AI is the technical present, and most definitely the future. “If an organisation thinks it does not need AI, that is today’s equivalent of saying we do not need emails,” warns McKelvey.
Companies, she adds, should be circumspect, but also optimistic in looking at the business opportunities and risks. But the question remains: can AI be trusted? ‘’Business leaders should always think carefully about trust and risk, says McKelvey. But today’s business challenge is not superintelligence. Today’s challenge is responsible implementation. We’ve already seen examples that remind us how powerful technologies need oversight, testing and governance. That’s why organisations need clear policies, human accountability and robust controls around AI use.”
The lesson isn’t ‘don’t use AI.’ The lesson is ‘use AI responsibly.’ She adds: “There is a lot that can be done from a technical architecture, governance and people training perspective to create safe and robust environments so that people can benefit from AI without leaving an organisation exposed.” Deloitte has deep expertise in supporting companies as they embrace AI.
[1] Michalski, J. (2025). AI ROI: The paradox of rising investment and elusive returns. tinyurl.com/3h3kb8xz.
[2] Deloitte. (2026). Deloitte CFO Survey Q2 2026: Growing AI optimism. tinyurl.com/mvaju5hx.
[3] Deloitte. (2026) GenAI Workforce Survey 2026. tinyurl.com/3p46mutv
[4] Deloitte. (2026). Navigate the economics of AI. tinyurl.com/au3vksch
