A majority of UK organizations are increasingly concerned about employees using unapproved software and AI tools, with 58% of cybersecurity decision makers identifying it as their top human-related cyber risk, according to a new report by KnowBe4.
The report titled “From Agentic Risk to Human Wins” has revealed that 55% of employees in United Kingdom have admitted to using unapproved AI tools, while 1 in 10 knowingly entered sensitive information into AI platforms despite understanding the risks.
According to the report, AI poses growing concern for British decision makers, highlighting that 42% of cybersecurity leaders identify AI-enabled attacks as a top driver of future human-related cyber security risks. Almost 49% of decision makers said that managing the safe use of AI tools and AI agents is one of their top concerns.
In fact, 46% of decision makers said they have specific targets for improving the safe use of AI agents in day-to-day workflows over the next 12 months.
The report disclosed that almost one in five (19%) of decision makers mentioned that AI tools or AI agents take actions autonomously in multiple workflows with limited human oversight. It also mentioned that worldwide, 58% employees have AI tools or AI agents taking actions autonomously in multiple workflows, including 17% with limited human oversight.
While the report also pointed out that 85% claimed that improvement is needed to ensure AI tools or agents operate within the organization’s security policies and approved risk limits.
“As organizations provide employees with LLMs and agentic AI tools to move faster, they may not always be putting relevant guardrails in place. Without these controls, an AI-enabled workforce will lead to greater risks instead,” the study noted.
“Almost half of employees themselves also admit that even when they know the safe action to take, time pressure or distractions could lead them to making a security mistake,” it stated.
The report stressed that agentic AI tools are now widely embedded in day-to-day work, with globally 58% of cybersecurity leaders reporting that agents are already taking actions within organizational workflows, stating that at the same time a lack of oversight and governance is leaving organizations exposed.
“Only 48% describe their AI use as formally approved and governed and 37% say while AI use is formally approved, governance is limited or unclear,” the report observed.
“Undeniably, AI tools and agents are reshaping the workplace, but organizations can’t afford to overlook the human element of cybersecurity,” said Javvad Malik, lead CISO advisor at KnowBe4.
“Our research shows that while UK businesses are embracing AI to drive productivity, many employees are still under pressure, using unapproved tools and regularly facing (and fearing) sophisticated threats such as deepfakes and phishing. Building a strong security culture, especially one that prioritizes education, behavioural support and safe AI adoption, will be critical to reducing human-related cyber risk in the years ahead,” he added.
The research features insights from 80 decision makers and 300 employees in the UK. Respondents represent organisations with 250 or more employees and span both private and public sectors among a wide range of industries such as information technology, healthcare, consumer services and others.
Also Read: Britain hit hardest by AI Job Cuts: Morgan Stanley
