Organizations Implement Multi-Pronged Strategy Against Unauthorized AI Risks

The use of officially unapproved artificial intelligence tools by employees in workplaces through personal accounts increases data security and legal liability risks for institutions. According to research, while a large portion of employees access unauthorized AI services from corporate devices, companies are developing comprehensive precaution strategies against this situation, including secure alternatives, data loss prevention software, employee training, and disciplinary processes in case of violations.

Verizon's Data Breach Investigations Report published in May reveals that 67 percent of employees access artificial intelligence services via personal accounts on corporate devices. Another analysis conducted by Harmonic Security in May showed that 65 percent of AI sessions conducted by employees through personal accounts were related to business activities, with these tools being used to write business emails, debug code, and conduct legal work. Furthermore, Harmonic's July report stated that every monitored organization had at least one employee in the first half of 2026 who sent data to a China-based artificial intelligence tool; this data included merger and acquisition plans, source code, financial projections, and legal materials.

Companies in the finance and technology sectors, such as BitGo and Docusign, use single sign-on (SSO) systems and data loss prevention (DLP) tools to ensure employees have access to approved and secure enterprise AI versions. While some risky models like China-based DeepSeek are blocked at the network level, non-governmental organizations such as JA Worldwide focus primarily on training sessions and secure routing systems hosted in private environments instead of costly monitoring tools due to budget constraints.