The Information reports large firms handling sensitive corporate work — including Palantir Technologies, NVIDIA and Booz Allen Hamilton — have begun demanding new safeguards or reducing or halting use of advanced models from Anthropic and OpenAI amid rising concern those models could learn from customers' IP. Microsoft is pushing its own products to capitalize on these security fears.

2026-09-14

The Information reports large firms handling sensitive corporate work — including Palantir Technologies, NVIDIA and Booz Allen Hamilton — have begun demanding new safeguards or reducing or halting use of advanced models from Anthropic and OpenAI amid rising concern those models could learn from customers' IP. Microsoft is pushing its own products to capitalize on these security fears.