Cloudflare said on the 7th that it launched an "identity-aware AI gateway" that helps corporations more clearly understand how workplace artificial intelligence (AI) is used.
As AI use grows within corporations, expense management and security concerns are also rising. The cost per individual AI request is falling, but as the volume of requests from employees and automation bots surges, billing expenses can jump sharply without prior warning. There is also the problem that even if sensitive information such as employee names and passwords or confidential files is sent to external AI providers, corporations may find it difficult to detect it immediately.
With this feature, for the first time you can identify which employee or AI agent sent a request to an AI model, set spending limits by individual or team, and automatically identify risky content. The accompanying "user insights" feature learns typical AI usage patterns for each user and automated system on the network and alerts the manager when activity deviates from those patterns.
Cloudflare said the feature can be applied without building a separate authentication system or changing existing work processes.
Max Baumgarten, a security engineer at Flexport, said, "When using shared API keys, it is hard to know who is using AI services, and it is practically impossible to apply existing access policies already in place for employees as-is," and added, "By integrating Cloudflare Access with the AI gateway, you can attach a verified identity to every request, and existing identity policies can be enforced directly at the gateway. Without building separate authentication systems for each AI tool, each team can safely adopt the AI services it needs."