According to a new report, The Grey Zone, from payroll and HR specialists, Zellis. a considerable gap or ‘grey zone’ exists where AI adoption is widespread, but impact, usage and confidence levels remain uneven across the workforce.
Zellis surveyed 500 UK employers and 100 employers in the UK and the Republic of Ireland (ROI), all at director level and above in companies with at least 1,000 employees, as well as employees in both the UK and ROI.
The report findings highlight that better AI alignment, adoption and integration at work represents an immense productivity and cost-saving opportunity. In the UK alone, organisations could unlock £40 billion a year in productivity gains, and a further £20 billion a year in operating-cost savings through better AI alignment.
When it comes to productivity, the research findings indicate that roughly 8% of current working time is wasted due to poor AI alignment. On a national scale, this equates to 1.7 billion working hours per year that could be redeployed to higher-value tasks. One in five leaders believe that closing the grey zone could cut operating costs by up to 10% and free up £1 in every £10 to be reinvested elsewhere.
Yet the benefits of aligned AI extend beyond economics. Nearly three-quarters of business leaders believe better alignment between leaders and employees, on how AI is used in the workplace, would reduce stress (73%), improve morale (73%) and improve trust between employees and leadership (72%). Whilst employees are less optimistic, the majority (60%) also believe better alignment would reduce stress, improve morale (49%) and boost trust (47%).
The majority of employees (75%) and leaders (74%) also agree that productivity would improve if AI was better aligned with how work is actually done. Additionally, 40% of employees stated they are more likely to stay with an employer that uses AI transparently.
The majority (67%) of all employees who use AI also agree it gives them back time for more meaningful or strategic work, indicating the role AI can play in supporting career progression, freeing time for more complex development driven tasks and training that push their personal and professional development.
The AI Disconnect
In the report, Zellis explicitly distinguishes between AI-led and AI-informed. The report notes that while leaders may intend for AI to be a tool for augmenting human decisions, employees often interpret AI-led as a lack of human oversight, particularly in sensitive areas like pay or recruitment.
Differences in perception when it comes to AI
When asked whether entering or checking basic data and records should be human-led or AI-led, around seven out of 10 (69%) employees chose AI-led, compared with only 44% of business leaders.
When It comes to recruiting or short-listing new employees, only 19% of employees said that these tasks should be AI-led, compared to 38% of business leaders. Similarly, just 8% of employees said deciding promotions or pay rises should be AI-led, compared to more than a third (35%) of leaders.
Additionally, although 63% of leaders believe they involve employees or teams in AI decisions, only 40% of employees who use AI agree they feel involved in decisions about how it is used. One in five (21%) employees disagree with the statement that their feedback about AI use in the workplace is valued and acted upon, although younger employees are most likely to agree that their feedback is valued, with 58% agreeing they feel heard.
Bridging the Gap
"When organisations are clear that AI is there to inform judgement rather than replace it, confidence grows," says Steve Elcock, Director of Product - AI at Zellis.
To bridge the gap, the report urges management to:
- Listen to everyday pain points: Focus AI on the repetitive tasks where employees want the most help.
- Co-design the strategy: Involve employees early and be explicit about which processes remain human-led to protect trust.
- Prioritise Transparency: Moving from "shadow AI" to open, collaborative implementation is now a primary driver of workplace culture and retention.
Abigail Vaughan, CEO of Zellis said: “This data makes clear this isn’t just about AI adoption, but about making sure AI is implemented with transparency and collaboration, and in ways that genuinely support how people experience work to reach their potential.”
She continued: “When leaders involve their people, communicate clearly and use AI to inform rather than replace human judgement, organisations can elevate their work and empower employees to be their best. There’s a huge value to be gained through AI when attention is paid to alignment.”
Management-Issues contacted Abigail to get her thoughts on what employees stand to gain from alignment and what can be done to ensure that AI efficiency isn’t just ploughed back into improving profit margins.
Q: Your report highlights how improved AI integration cuts operating costs. How can leadership invest this saving back into employee development?
A: As with any efficiency improvement, employers will have options regarding how to realise the benefits which will range from margin improvement and return to shareholders through to re-investment to accelerate growth. One choice for reinvestment will be ensuring that all employees are AI literate and have skills for the modern workplace, with no one left behind. Even if an employer cannot guarantee employment, it can commit to ensuring that people have the skills to be employable. This is a societal imperative.
Q: How can organisations avoid a 'productivity trap' and instead use AI to 'lift people up' as Steve Elcock suggests in the report?
A. To avoid the ‘productivity trap’, leaders need to expand the definition of productivity. It’s not output per hour; it’s the value created per human hour. AI can create extraordinary speed and accuracy, but it must be paired with human judgement, empathy, creativity and leadership because those are the multipliers.
My colleague Steve Elcock talks about using AI to ‘lift people up’, and that’s exactly the mindset needed - AI should elevate roles, not compress them. That means deliberately designing systems where AI frees people up to move closer to customers, decisions and innovation.
Without intentional governance, organisations risk using AI to push people harder instead of helping them work smarter. The companies that get this right will be those that embed AI into a broader cultural shift, one where humans remain at the centre of how value is created.
Q: You've noted that employees would be more likely to stay with a transparent employer. This being the case, does management have a responsibility to be transparent about how those savings are being used? Can we expect to see increased funding on the AI upskilling and training that 74% of leaders say is now essential?
Most businesses I work with are making early investments in AI upskilling, tooling, and ways of working, ahead of seeing the full benefits.
Generally, as leaders, we are not transparent enough about the investments we make - many of which have a very positive impact on jobs and wellbeing. Thoughtful investment strengthens both the business and its people. People rightly want to know how AI will affect their roles, their development and their future, and having an open dialogue is the best way for people to feel more valued and not more vulnerable.
