Article summary: A new BCG study reveals that 54% of employees would use AI at work without approval, exposing companies to rising security risks. Only 36% feel adequately trained, and just 13% report AI agents being integrated into workflows. To address the ‘Shadow AI’ trend, the report urges employers to upskill staff, redesign workflows, experiment with AI agents, and establish strategic guidance to ensure responsible and effective AI use.
A new global study by Boston Consulting Group (BCG), involving survey responses from over 10,600 workers, reveals a growing trend of ‘Shadow AI’ use in the workplace. More than half of respondents (54%) say they would use AI tools without formal approval – a behaviour especially common among Gen Z and Millennial employees.
Compounding this risk, only a third (36%) of employees feel adequately trained to use AI, and just 25% of frontline workers say their managers provide sufficient guidance.
This combination of unregulated AI use and insufficient support leaves employers vulnerable to security threats – including cyberattacks, data breaches and compliance issues.
The research also found that UK employees are saving time through AI use, but many lack clear direction on where to reinvest it. Just under half (49%) claim they save more than an hour each day, but only 40% receive guidance on where to spend this extra time, suggesting missed opportunities for productivity and culture.
Other key findings from the report:
- AI is well and truly mainstream: 72% of global respondents use AI regularly at work.
- Global South leads AI adoption: India at 92%, the Middle East at 87%, and Brazil at 76%.
- Job-loss fears persist: 41% of global respondents are worried their roles could disappear within the next decade (with high-use regions reporting the greatest fears).
- The UK leads on workflow redesigns: 58% of UK respondents say their company is rethinking workflows (compared with the global average of 50%).
- Still early days for AI agents: While 75% believe AI agents will be critical to business success, only 13% say these tools are being integrated into workflows.
What does this mean for HR leaders?
Experts are increasingly alarmed by employers’ lack of strategic response to growing AI issues. “From my conversations with businesses across sectors, it is concerning how many HR professionals have not yet considered the practical risks of unregulated AI use,” says Lorraine Hunt, Founder of AI Automation Dept. “It’s often the small overlooked things that could lead to major data breaches and reputational headaches. These are not theoretical risks – I have seen and heard of some very poor practices.”
HR teams should forge strong partnerships with IT to develop comprehensive, accessible AI rules and guidance, Hunt advises. “The idea is not to block AI, but to embrace it with clear, strategic intention.”
Those businesses without a thoughtful AI strategy will not only be open to risks, they’ll also see little productive impact. “Companies cannot simply roll out GenAI tools and expect transformation,” said Sylvain Duranton, Global Leader of BCG X and a coauthor of the report. “Our research shows the real returns come when businesses invest in upskilling their people, redesign how work gets done, and align leadership around AI strategy.”
“The idea is not to block AI, but to embrace it with clear, strategic intention.”
Addressing the Shadow AI trend: Four next steps
For those looking to address the shadow AI trend and transition from tools to meaningful impact, the report outlines four key actions.
1. Commit to comprehensive AI upskilling for your people: Think beyond offering a ‘tick-box’ digital course for AI beginners. Allocate appropriate levels of investment, time, and leadership support to AI upskilling.
2. Measure AI impact: Keep a close eye on how your AI tools are improving metrics such as productivity, quality, and employee engagement.
3. Reimagine workflows to prepare for the future: Empower and equip your employees to integrate AI into their workflows. Additionally, carve out time to consider AI’s impact on work through the lenses of both individual employees and the wider workforce. From there, make a plan to support workforce deployment via upskilling and reskilling opportunities.
4. Get going with AI agents: Experiment rigorously with agents to accelerate the experience curve, assessing both the impact and potential risks along the way.
Don’t overlook AI’s impact on workplace culture
Alongside these four actions, employers shouldn’t ignore the implications on company culture, especially as agentic AI becomes mainstream.
“Looking ahead, as AI agents enter the workplace and smarter tools are released, it is only going to raise the stakes,” Hunt adds. “I would urge HR leaders to take the lead in creating a workplace culture where responsible AI use is the norm – where employees feel confident to innovate, but understand the boundaries that are in place.”
