As Kate Dodsworth, Partner at Shoosmiths, explains, “while the benefits of AI are widely publicised, the legal and practical consequences of AI in the employment context are only just beginning to emerge. Unintended consequences can be uneven practice, heightened risk and avoidable pressure on both employees and people teams.”
The real trade-off is rarely ‘productivity versus wellbeing’. It is productivity versus capability: whether AI adoption strengthens human judgement and skills, or gradually replaces them with work intensification, skills atrophy and a loss of professional identity. HR’s role is to make that choice explicit — and to build the governance, skills and work design that keep performance and people outcomes aligned.
Given knowledge workers form a significant part of the UK workforce, this article focuses on knowledge-based roles. We focus on what people professionals can do now, to shape AI use, in ways that improve outcomes for people and the organisation.
Work intensification and ‘work creep’
UC Berkeley research points to three common effects of AI usage: task expansion, blurred boundaries between work and home, and increased multitasking. The result can be longer working days and wider role scope. While this may look like higher discretionary effort, it is often unsustainable. Over time, intensity can reduce focus, increase errors and weaken decision quality — undermining the productivity gains AI is meant to deliver.
Cognitive fatigue (‘brain fry’)
AI ‘Brain fry’ has been described as “mental fatigue from excessive use or oversight of AI tools beyond a person’s cognitive capacity”. It differs from burnout; it is driven by constant monitoring, checking and accountability for AI outputs. The risks are practical — decision fatigue, higher error rates and increased intentions to quit. People functions are among the professional groups reporting higher levels of this strain, which should prompt HR teams to model sustainable use and set realistic expectations for oversight.
Skills erosion and ‘cognitive offloading’
Early research suggests that consistent AI use in specialist domains can reduce autonomy and resilience, with signs of skills erosion after relatively short periods. For employers, the risk is not only individual performance but organisational capability; if learning and judgement shift from people to tools, expertise becomes harder to build and easier to lose. People professionals can counter this through work design, learning pathways and clear expectations about when employees should rely on tools versus practise core skills.
Job insecurity and shifts in professional identity
Some professionals report feeling ‘deprofessionalised’ as work moves from expert practice to supervising AI outputs. Where AI takes over tasks connected to purpose and craft, employees can feel reduced to passive overseers — affecting motivation and retention. There are also disproportionate impacts: evidence suggests entry-level roles can be more exposed to displacement, with knock-on effects for early career development and the future skills pipeline. HR should treat this as a workforce planning issue, not only a change management challenge.
What AI changes for HR teams: risk, workload and credibility
Misuse and ‘over-production’ of workplace issues
Easy access to generative AI can lead to longer, more complex grievances and correspondence, sometimes referencing irrelevant or inaccurate points. That can increase the time needed to investigate, respond and resolve concerns — while raising the risk of misunderstanding and escalation. The result is a hidden workload for people teams and managers, and potentially higher legal costs. HR teams will need clear guidance on appropriate AI use and a process that cuts through volume to get to the underlying issue quickly and fairly.
Confidentiality, data protection and information security
Most employers understand the risks of sharing personal data and commercially sensitive information. In practice, risk often enters through routine use of publicly available tools: an employee pastes identifiable information into a prompt, uploads documents, or asks a tool to rewrite correspondence that includes confidential details. HR can reduce exposure by setting clear ‘do not share’ rules, providing approved tools, and building data protection awareness into everyday guidance — not only annual training.
What future-focused HR teams should do now
AI can reduce effort on repetitive tasks and improve access to information. But the value comes from deliberate design and clarity on where AI supports work, where it should not be used, and how to protect fairness, capability and trust. CIPD guidance is clear that wherever AI has a human impact, people teams need to be as involved as IT, legal and compliance.
1) Put governance in place before scale
Start with a small number of shared principles and minimum standards. This is less about restricting innovation and more about creating consistency, psychological safety and accountability. HR should co-own these AI practices and ensure they are reflected in policies, processes and manager practice. AI principles may include:
- Approved tools and what ‘safe to use’ means (including data handling and prompt guidance).
- Where AI must not be used (for example, highly sensitive data, formal employee relations documentation, or regulated decisions).
- Human oversight requirements (‘human-in-the-loop’) for higher-risk decisions.
- How outputs should be checked, documented and challenged.
- Consequences of unauthorised or inappropriate use, aligned to disciplinary policies.
2) Take a people-centred approach to implementation
The InnovateUK BridgeAI and CIPD’s people-centred AI work underline a simple point: implementation is not an IT rollout. HR should help leaders define the problem to solve, test use cases, assess impacts on roles and skills, and iterate.
3) Question the efficacy of digital tools
Ask two questions of every tool and use case: does it improve work outcomes, and is it reliable enough to be trusted for that task? Research shows that AI efficacy is linked to increased productivity, engagement and job satisfaction.
4) Invest in skills and workforce planning
Workforce readiness remains a barrier to successful AI adoption. Research suggests many organisations believe fewer than a third of their people have the digital skills needed to operate effectively. HR can address this through targeted AI literacy (including critical thinking and data handling), role-based upskilling and strategic workforce planning to protect the pipeline of expertise.
5) Set expectations for sustainable use
If AI increases pace and multitasking, HR should treat sustainability as a design requirement. Set norms for when tools should be used, how much checking is reasonable, and when teams should slow down. Look for warning signs — rising out-of-hours work, decision fatigue, and growing time spent validating outputs. Encourage employees to think about how digital tools can optimise their outputs, rather than encouraging them to tinker with the technology — this often leads to minimal or shallow benefits. People teams can lead by example, using AI to remove low-value work while protecting time for judgement, coaching and relationship-based practice.
6) Protect fairness in employee relations and high-stakes decisions
In employee relations, volume and complexity can increase while accuracy stays uneven. Keep high-stakes decisions ‘human-led’. Shoosmiths advise, “AI may support summarising information or checking consistency, but it should not be the sole basis for outcomes such as disciplinary warnings, dismissal, or adjustments that could have discriminatory effects. Where AI-generated grievances appear, focus less on the text and more on clarifying the underlying concern through conversation and evidence, so the process remains fair and proportionate".
AI can help organisations do more but future readiness depends on whether we strengthen or weaken human capability in the process. HR’s opportunity is to move the conversation from ‘tool adoption’ to adapting work design, bringing governance, skills and sustainable performance to the forefront — aligning the people, culture and tasks to achieve the productivity gains that AI has promised to deliver.
Further reading
How people professionals can develop, deploy and use AI in an ethical, legal and sustainable way | CIPD
AI rollout ideas for people professionals | CIPD
How organisations can take a people-centred approach to AI | CIPD
AI in the workplace | CIPD
Strategic workforce planning | CIPD
AI governance report | Shoosmiths
Ethical AI: Turning intent into impact
Unintended Consequences of Artificial Intelligence in Employment
AI is reshaping workplace grievances
