Lindsay Gallard, Chief People Officer at Six Degrees offers a note of warning to HR as technology enters the workplace.
Encouraging employees to embrace AI in their working lives may be easier than initially expected. According to recent research from Deloitte, around 7 million people claimed they were using GenAI for work in 2024, a substantial increase from 4 million in 2023 – indicating that employees are by no means universally opposed to using AI in their jobs as is sometimes suggested. In practice, it seems staff are willing to take on emerging technology and learn new skills if it helps them perform their daily tasks more easily. Consequently, many have been quick to recognise the benefits GenAI can bring, especially for brainstorming and information gathering.
On the one hand, this sounds extremely positive, paving the way for the introduction of more AI tools and innovative ways of working. However, on the downside, the research also highlighted a number of risks associated with this growing trend that may not be immediately apparent. It pointed out that the majority of those using GenAI for work purposes may be doing so without official endorsement by their employer. What’s more, it went on to say nine in ten organisations lack a policy on whether the use of Gen AI is acceptable, or have a governance structure if it is.
Given the problems of bias, unreliable results and AI ‘hallucinations’, this should be raising multiple red flags at senior management and board level. Left unaddressed, there’s serious risk of poor, even dangerous outcomes that could at the very least be damaging reputationally and financially, and in the worst case possibly ruinous.
What constitutes inappropriate use of AI?
If this doesn’t already sound like a minefield for unsuspecting businesses, there’s another twist in the tale concerning the misuse of tools that have already been sanctioned. Often, applications that have been given the go-ahead as fit-for-purpose by IT have unwittingly bypassed approval from HR and legal teams. This begs the question whether anyone has investigated if these tools could be deployed inappropriately, how this might manifest itself, and what would be the consequences?
Take for example AI-powered sales management applications for automated note-taking and forecast analysis. While these tools are designed to improve performance and optimise sales processes, they could have data privacy loopholes. Many have powerful, sophisticated search capabilities, enabling sales leaders to retrieve information about how deals are being handled across multiple applications, including transcripts from calls and online meetings. Taking into consideration that sales staff are often home-based, frequently working remotely without direct oversight, the temptation for their managers to use these applications to scrutinise an individual’s behaviour and aptitude might be hard to resist. If this data is analysed with the intention of appraising performance and re-setting goals, then it could be deemed as misuse of personal data and an invasion of privacy. At worst, it might even be viewed as harassment or bullying.
Governance must incorporate staff wellbeing
Without proper governance, anyone who has access to these tools could run similar searches for dubious reasons across their teams and colleagues. To minimise the chance of any kind of misuse, organisations must ensure they have well thought out selection and implementation processes for all AI deployments, no matter how harmless they might appear at first glance. This should include an appropriate level of consultation with legal and HR departments before procurement to understand the possible impact on employee wellbeing and to guard against unauthorised use. As a minimum, organisations should consider the following:
- Evaluation Criteria
Documenting the evaluation process to cover more than simply technology deliverables. Also, addressing how implementation might impact privacy and confidentiality standards for users and their staff. Checking how authorisation levels are managed, and what safeguards are in place to prevent misuse. Plus, challenging whether security controls are robust enough.
- Trial Periods
Also having mandatory trial periods before committing to purchases, and encouraging the set up of user groups to thoroughly road test a new product or upgrade, and report on any concerns. Avoid rushing through this stage as feedback gained at this point will help inform decision-making and make for a smoother roll-out, if the trial is positive.
- Legal Risks
In addition to HR, ensure legal and compliance teams are involved to advise on risks and how to mitigate them. Take onboard their input, and don’t view their contribution as simply a box-ticking exercise.
- Employee Communication
Always keep employees in the loop about what’s happening. Have a variety of representatives from different areas on the user group and ensure any concerns are addressed. Explain why tools are being trialled and rolled out, and how employee rights will be protected.
How HR can guide company-wide AI evaluation and deployment
While much of this is standard project management, it’s crucial that HR are involved in the process. Input and feedback from staff across an organisation can be subtly different when requested by HR professionals. Appreciating these nuances and providing staff with a safe environment to voice their concerns will help organisations leverage the value of AI more effectively.
Going forward, HR must continue to play their part in the adoption of any new technology. With the forthcoming Employment Rights Bill strengthening protection against unfair treatment at work, employers will need to think carefully about how, or if, they use technology to monitor performance.
Then, there’s the question of how HR teams evaluate the quality of work produced by employees that are using AI, if it is sanctioned for use. Keeping tabs on what’s computer-generated and what’s not is going to require revised thinking around the evaluation of work and career progression. It stands to reason that if staff are already using GenAI to help with their daily tasks, they are no doubt using it to assist in completing their appraisal forms too. This will make it harder for HR to maintain objectivity in the performance assessment process.
Additionally, the introduction of day-one rights for flexible working along with protection against unfair dismissal will make managing probationary periods more complex. As remote and flexible working arrangements can be in place before a formal one-month appraisal, there’s considerable pressure on employers to ensure competencies are being assessed fairly and consistently from the day a new recruit joins. Judicious use of AI for monitoring might be the answer, provided its use is always transparent.
When all’s said and done, guiding managers and staff in the appropriate use of AI-enabled tools in a manner that respects employees rights clearly belongs within the remit of HR.
I believe keeping the human emphasis on people, communication, and collaboration will help to prevent AI misuse, encourage adoption, and drive better outcomes for both staff and employers.
