HR professionals welcome the opportunities presented by AI and are moving forward with implementation ahead of many other teams across the business, but are struggling with challenges within their organisations.

A report released today reveals there is a lack of understanding of how AI can address business process issues, a siloed approach to AI within organisations, and a lack of confidence in data quality and human expertise to deliver good AI outputs. The data* combines more than 6,000 senior executive voices from a cross-section of British businesses. It shows an overwhelming 96% of HR professionals say they are already using or are planning to use AI to enhance business processes.

Obstacle One: A lack of skills and understanding

The study also shows 39% have successfully implemented an AI-related project, which is more than almost all of the other functions surveyed. Of those HR professionals who have tried unsuccessfully to implement AI use cases, more than one-third (36%) said this was because they did not have the right skills and experience to manage the project.

AI adoption requires a variety of skillsets. Employees in organisations will need to be upskilled on AI tooling and how to get the most out of AI products – for example effective prompt engineering. In addition, successful organisations will ensure a cross-functional skillset is involved in AI use case selection and implementation, including project management, financial, organisational change management, data privacy and protection, cyber security, ethics, user experience, technology, data and workflow skills are represented.

The right critical thinking in identification of AI use cases is at the heart of other challenges, as almost half (48%) say they did not really understand the business problem they thought they were addressing with AI.  To truly unlock the real potential of AI businesses must start with the end in mind and understand the problem they are looking to solve and value it will deliver, before deciding which technology solution can best solve the problem.

Obstacle Two: A siloed approach to AI

The report demonstrates a siloed approach to AI is undermining progress.  Less than one-quarter – (23%) of HR professionals – say their organisation has put together a cross-functional committee to consider all of the governance, data protection, and end-user best practices.

Instead, 29% say that different functions within their organisation procure their own software, make their own policy, and have implemented their own guardrails and best practice. 26% say certain teams and departments are using AI tools independently, without a cohesive, company-wide strategy.

The highest performing teams operate cross functionally where complementary skills come together to solve business problems.  The winning organisations will be those who take a holistic approach to the use of AI and the interconnectivity of its use cases.

Obstacle Three: Data quality and expertise

The third major stumbling block relates to data. Of those who have not been successful with AI implementation, 36% said this was because they had not sufficiently reviewed and cleaned up their data before embarking on the project.

29% put lack of success down to underestimating the level of data expertise and resource required and 22% admit their data is not organised or accurate enough to produce reliable AI outputs. The report also finds more than one in five (21%) say their organisation has not secured the required permissions or maintained a robust audit trail to ensure data integrity.

Operational excellence and rigour is required around data governance.  Data used to train models or that will be surfaced by AI needs to be high quality, relevant, up-to-date, complete, correctly tagged and free from bias. Getting this right will only accelerate the success of AI or data led projects. Getting it wrong can cause issues, unreliability or bias with outputs and lead to faulty decision-making.

Katie Obi, Chief People Officer, OneAdvanced said: “Whilst HR is among those functions that have been most successful with implementing AI projects so far, the Report* reveals a lack of understanding of how to implement and use AI effectively. HR professionals have a responsibility to improve their own data skills and critical thinking around the use of AI, acting as pioneers for their organisations, whilst also equipping their organisations with the right skills for AI adoption and critical thinking.

“AI outcomes are based on data, and if the data going in is poor quality, or has inherent bias, then the information coming out will be inaccurate and any decisions based on it will be flawed. HR professionals are excellent collaborators with a strong knowledge of processes, ethics, and human behaviour, used to pulling together cross-functional teams and working effectively with people across the entire organisation. We can combine these skills with critical thinking, as well as our adeptness with legislation and compliance issues, to help guide our organisations towards more effective AI implementation. When done well, this would drive improved business, customer, and employee experience outcomes, reduce bias, and increase productivity and operational excellence.”

*The OneAdvanced 2024/25 Trends Report