Dashlane survey finds widespread workplace credential sharing with AI tools

src-1b4567a

Nearly one in five UK workers surveyed by Dashlane said they had shared a password, API key or other login credential with an AI tool or agent, highlighting a growing identity-management challenge for enterprise security teams. The findings also point to a gap between employees’ use of AI and security leaders’ confidence that they can see all such activity.

Dashlane, a credential security company, surveyed 1,000 UK workers and 250 IT executives for its report, Misplaced Trust: Inside the Workplace’s New AI Security Blind Spot. The UK research formed part of a wider survey of 3,800 respondents across the UK, US and France. The results were published on November 19, 2026.

Employees are extending access to AI

In the UK worker findings, 19% reported sharing a password, API key or login credential with an AI tool or agent. Nearly half, 48%, said they had allowed AI to carry out at least one activity without first seeking permission. Examples included access to email, company files and customer information.

Other responses suggest that employees may grant or retain access without fully considering its scope. Some 15% said they had given an AI tool access to a file, system or account and later discovered it should not have had that access. Among that group, 45% said the access was still active. Separately, 37% said they would be willing to give their login credentials to a new AI tool after using it for one week.

The survey also found that 23% of workers had copied sensitive information into AI that they would not normally share with a colleague, website or other work system. One in five, 21%, said they had not checked where information would be stored or sent. Only 25% said they always consider security risks when using an AI agent at work.

Visibility and confidence do not align

Dashlane’s executive responses suggest that AI-related access has already become an incident-management concern. Seventy-two percent of the surveyed IT and security executives said their organization had experienced a security incident involving an AI agent’s credentials or access during the previous 12 months.

At the same time, 92% of executives said they were confident they knew about every AI agent operating in their organization. That confidence sits alongside the employee finding that 25% use AI tools their employer does not know about. The figures do not establish the precise cause or nature of the reported incidents, but they underline the difficulty of managing identities and permissions when AI use is not fully visible to IT.

Business pressure is another factor. More than two-thirds, or 68%, of UK IT security executives said they felt caught between pressure to deploy AI quickly and the need to keep their organization secure. Dashlane Chief Product Officer Christophe Frenet described the pattern as an “AI Colleague Effect,” in which users may trust AI systems with credentials and permissions without understanding what those grants enable.

Apply identity controls to AI agents

For security teams, the reported practices raise familiar questions about credential exposure, excessive privileges and access that remains enabled after it is no longer needed. If credentials or sensitive company information are provided to external AI services, organizations may also need to understand how those services handle and store the material. The survey does not identify specific AI products or document a particular incident involving data exposure.

Frenet said established identity-security principles should apply to AI systems as well as human users, including just-in-time provisioning, least privilege, credential hygiene and authentication. In practical terms, organizations can use those principles to set rules for when agents receive access, limit permissions to necessary tasks, and review or revoke access when it is no longer required.

The findings also support treating AI-agent credentials as part of an organization’s broader credential-management process, rather than allowing employees to share personal or work logins informally. Security leaders may need processes to identify approved agents, assess the data and systems they can reach, and investigate access when an incident occurs. Dashlane’s survey does not prescribe a specific product or control set, but its results show why visibility into AI use and accountable access management are becoming linked concerns.