The updated EU AI Act came into force yesterday, Sunday 2nd August. WalkMe, an SAP company, explains why this could be the moment shadow AI really starts to hurt businesses.

Article 50 introduces new transparency obligations – in some instances requiring organisations to disclose when text has been artificially generated, and informing people when they are interacting with an AI system. Failure to comply could result in being fined €15 million, or 3 % of total worldwide annual turnover.

Tal Carmi, WalkMe’s CIO, warns that shadow AI will make it difficult for businesses to meet these new expectations. At a top level, many organisations lack visibility over how employees are using AI – preventing them from governing activity, or demonstrating responsible practice. Global research from WalkMe found:

  • 45% of employees used unsanctioned AI tools in the previous 30 days
  • 36% used unsanctioned AI tools with confidential data
  • 34% did not know which AI tools their employer had approved
  • Only 21% had been warned about their employer’s AI policies

UK companies that operate in the EU, serve EU customers or purchase AI services from EU providers, need to meet the new rules. But as the research shows, many of them don’t know how their employees are using AI.

The comment below outlines how organisations can act – getting visibility and helping employees make the most of the right AI tools. I’ve also got the full research report, if you’d like to see it? Just let me know – otherwise, I hope the below is useful for a story.

Tal Carmi, CIO at WalkMe, says: “The updated EU AI Act means businesses are now in a race against time to tackle shadow AI. This is already a widespread problem, with our research showing that almost half of employees use unsanctioned tools at work. With shadow AI creating a more serious governance, risk and compliance challenge than ever, businesses need to act now or face the consequences. Leaders need visibility into which AI tools employees are using, what information is being shared and how people are relying on AI in their day-to-day work. Without that understanding, they cannot govern AI use effectively or demonstrate responsible practice.

“But leaders need to go further than simply addressing the use of unapproved tools. They must understand why employees are turning to them in the first place. Shadow AI is often a symptom of a wider adoption problem: employees will bypass sanctioned tools if they are difficult to use, do not fit naturally into existing workflows or fail to help them get work done. Businesses need to provide the right tools and support employees at the right moment, embed guidance into everyday workflows and remove the friction that drives them elsewhere. This will give organisations greater visibility and control, encourage adoption of approved tools and ultimately improve productivity. Strong governance depends on making approved AI the easiest and most effective choice for employees.”