AI and inclusion: why human judgement still matters
A recent lawsuit against Meta has raised uncomfortable questions about the growing role of artificial intelligence (AI) in people management and workplace decision-making. According to allegations reported by The Guardian, a group of employees claim that AI-powered decision systems were used by Meta to identify individuals for large-scale layoffs. Workers on maternity leave, parental leave and disability-related absences were disproportionately affected. Most concerningly, perhaps, the 71-page complaint states that, “Meta did not assemble the termination list through the considered judgment of managers who knew the work.”
Rather, a combination of AI performance ratings, keystroke and activity-monitoring metrics based on AI usage scores were allegedly used to make decisions. Meta disputes the allegations and maintains that workforce decisions were made by people, not AI. Nevertheless, the lawsuit raises urgent concerns about bias, trust, equity, inclusion and fairness in the workplace, that those of us who work for employee inclusion and wellbeing must heed.
AI: replacing or revitalising managerial judgement?
At first glance, AI appears to offer exactly what organisations need – the promise of greater efficiency, higher productivity, faster decision-making, greater consistency and improved performance. For organisations and managers facing increasing complexity, this is an attractive proposition.
AI can clearly contribute positively to workplace inclusion. It can identify disparities in promotion rates, flag potentially biased language in recruitment materials, improve accessibility through translation and speech-to-text technologies, undertake sentiment analysis, and help organisations understand patterns in employee experience data. As a support tool, AI has enormous potential.
The problem is when we assume that ‘objective’ data-driven decisions are inherently superior to ‘subjective’ judgment-based decisions and therefore automatically fair or inclusive.
The line manager as linchpin
At IES, our research consistently finds that line managers play a critical role in shaping employee experience. This is particularly true in areas such as maternity and pregnancy, parental leave, disability, neurodiversity, working carers, long-term health conditions, wellbeing and workplace adjustments.
Policies matter. Organisational culture matters. But employees often experience those policies through everyday interactions with their manager. Employees don’t simply want accurate decisions. They want to feel seen, heard, understood and respected. They want to know that somebody has genuinely considered their circumstances rather than accepted a recommendation generated by an automated system that cannot appreciate the complexity of their experience.
A supportive manager can make the difference between inclusion and exclusion, between retention and resignation, between psychological safety and disengagement. As AI becomes more embedded in workplace processes, employers face a choice about how this technology will shape the manager-employee relationship.
Future Scenario 1: AI replaces managerial judgment
In this world, management decisions become increasingly influenced by algorithmic recommendations, performance scores, predictive models and productivity metrics. Managers become administrators of automated outputs rather than decision-makers in their own right. AI systems learn from existing data and historical organisational patterns. If these patterns reflect inequalities, hidden assumptions, and unconscious biases, AI will replicate these at scale. Inclusive leadership, on the other hand, entails challenging and questioning precisely these historical decisions and past patterns of recruitment, promotion, and progression that disadvantage minority groups. CIPD’s research on AI in the workplace argues that AI adoption should be viewed not as a technology project but as a people and organisational change challenge instead.
Future Scenario 2: AI enhances managerial judgement
This is the more optimistic (and my preferred) possibility. Rather than replacing line managers, AI could help them become better managers. Research by the Chartered Management Institute (CMI) has found that 82% of people who move into management roles receive little or no formal leadership training, creating a generation of “accidental managers” imposing high costs on organisations and the wider UK economy.
Staff decisions calling for managerial judgement depend on context. More importantly, they depend on relationships. A good manager sees the whole person. They understand the circumstances shaping performance and apply judgement, discretion and care when making decisions.
But managers are increasingly responsible for large teams and navigating complex workplace issues. AI could provide immediate access to policies, evidence-based guidance and practical resources that help them support employees more effectively. Gartner research reports that 45% of managers found that AI adoption helped them improve their teams’ work.
Imagine a manager supporting a neurodivergent employee. Instead of searching through lengthy policy documents, they receive tailored guidance on potential workplace adjustments and inclusive communication approaches.
Imagine a manager preparing for a conversation with someone returning from maternity leave. AI could provide prompts, resources and practical suggestions to ensure those conversations are meaningful and supportive.
Imagine managers receiving AI guidance on disability accommodations, flexible working requests or wellbeing concerns in real time.
In these scenarios, AI does not make the decisions. It boosts managerial capability. It becomes the coach, the adviser and the knowledge resource that helps managers exercise better judgement rather than replacing it. IES’s study of AI use in the Restart employment support delivery confirms that .
The human advantage we risk forgetting
The most significant advances in workplace inclusion have occurred because managers saw possibilities that existing systems overlooked. They hired people from unconventional backgrounds, championed flexible working, supported successful return-to-work, created reasonable workplace adjustments, challenged exclusionary practices, and built environments where people felt able to contribute fully and authentically.
The lesson from the ongoing Meta case is not that organisations should reject AI. Used responsibly, AI can improve accessibility, strengthen inclusion capability and help managers navigate increasingly complex employment challenges. But when it comes to supporting employees with sensitive needs and marginalised identities, algorithms cannot replace trust, empathy and human understanding.
Inclusion is ultimately a human endeavour: the organisations that succeed will not be those that use AI to replace human judgement, but those that use it to strengthen managers’ ability to recognise potential, navigate complexity and create truly inclusive workplaces.
Questions to consider
If your organisation is navigating the intersection of AI, management and inclusion, consider these three questions:
- Are your AI tools improving managerial judgement or replacing it?
- Are your DEI efforts resilient enough to withstand changing organisational priorities and external pressures? Watch my webinar on What is Really Sustaining (or Undermining) DEI in UK Workplaces.
- Do you have the evidence, data and organisational insight needed to build genuinely inclusive workplaces? Explore the IES DEI offer and contact me at: [email protected]
Any views expressed are those of the author and not necessarily those of the Institute as a whole.