Workers Use Personal Accounts for Nearly Half of Enterprise AI Chats

Generative AI has spread rapidly through corporate workflows, but employees who use unapproved tools or accounts create a layer of “shadow AI” outside company oversight. The practice can expose customer information, internal documents or source code to third-party services, while weak identity controls make access difficult to audit. Data entered through consumer accounts may also be retained or used to improve publicly available models, raising security, privacy and compliance concerns.

Akamai said its latest research found that nearly half of employees’ enterprise AI conversations were conducted through personal accounts, leaving a substantial share of activity beyond corporate identity and permission systems. Security specialists recommend that companies adopt identity federation measures such as single sign-on, or SSO, to centralize access policies, manage account lifecycles and preserve usage records. The findings underscore a widening governance gap as workplace AI adoption outpaces formal controls.