19% of UK employees gave a password or key to AI, Dashlane reports
72% of IT executives report an incident with the access of an AI agent, but 92% are sure they know all agents.
Dashlane, a company that keeps login data safe, did a study of 1,000 UK employees and 250 IT executives. The study has 3,800 persons in the UK, the US and France. Dashlane reports that 19% of the UK employees gave a password, an API key (a code that lets software connect to a system) or a login to an AI tool. Also, 48% let AI do a task with no permission step.
What employees do
- 37% are prepared to give their login to a new AI tool after one week.
- 23% put data in AI that they do not give to a colleague.
- 21% did not check where the data goes.
- Only 25% always think about risk when they use an AI agent.
The proof and the limits
15% of employees gave an AI tool access to a file or an account that it must not use. Of these, 45% report that the access stays open. 72% of IT executives report an incident in the last 12 months. But 92% are sure they know all agents. The report gives no count of the incidents.
What Dashlane advises
Dashlane reports that the rules for access by a person must also apply to AI agents. It names the minimum access and good control of logins.
Insight: Agents with a person’s login hide in the person’s name
Dashlane finds that employees give logins to AI. Nudge Security reports that agents made by staff use the access of the person who made it, and logs show the employee, not the agent. ServiceNow found that only 23% of companies have one owner for agent rules. This reads as a blind spot: agent steps look the same as staff steps, and the 92% can count only the agents that they find. Leaders must count access, not tools.
Sources
- A Fifth of UK Employees Share Passwords with AIinfosecurity-magazine.com
Posted