Companies including Goldman Sachs, Cisco and JPMorgan Chase are tightening employee access to AI tools to manage costs, security risks and productivity, according to a November 11 report by Business Insider. The changes affect access to advanced AI models and coding assistants as organizations reconsider unrestricted employee usage.

According to the report, Cisco plans to introduce token budgets for managers after allowing coding teams months of unrestricted use. JPMorgan has capped some engineers’ Claude Code spending at $2,000 per month, while Goldman Sachs is taking a more targeted approach to access to advanced models. 

While the companies are seeking to link AI usage more closely to productivity, quality and business value, the shift also reflects concerns about unauthorized usage and data exposure. 

A 2025 KPMG survey cited in the report found that 44% of more than 1,000 US workers had used AI at work without authorization or in inappropriate ways, while 46% had uploaded sensitive company information or intellectual property to public AI platforms.

Establishing Controls for Enterprise AI Access

For CDOs, tighter access policies raise questions about which employees can use specific AI models, what enterprise data those tools can access and how usage is monitored. Data leaders should work with security, IT and business teams to establish model permissions, approved-use policies and monitoring processes.

These controls must also address shadow AI, where employees adopt tools outside approved enterprise systems. AI governance roles and responsibilities need to be decided to establish accountability as AI adoption expands.

AI access is increasingly becoming a governed enterprise resource. Clear permissions and monitoring can help organizations balance adoption with cost discipline, data protection and accountability.