Home / Technology / AI Data Fears: Firms Block Sensitive Use
AI Data Fears: Firms Block Sensitive Use
14 Sep
Summary
- Concerns rise over AI models retaining sensitive corporate data.
- Palantir seeks zero data retention guarantees from Anthropic.
- Nvidia and Booz Allen limit Anthropic's AI for proprietary tasks.
Concerns are mounting over the potential for artificial intelligence models to retain or extract sensitive corporate information. As of September 14, 2026, major technology and consulting firms are implementing safeguards.
Palantir is reportedly pressing AI provider Anthropic for assurances of zero data retention. This demand precedes making Anthropic's models available through Palantir's software platforms.
Meanwhile, Nvidia has confined Anthropic's AI models to less sensitive internal tasks. Nvidia is prioritizing its own Nemotron models for any proprietary projects requiring advanced AI capabilities.
Booz Allen has also taken a cautionary stance. The firm has issued directives forbidding its staff from utilizing Anthropic's commercial model on cybersecurity projects that handle proprietary data, underscoring broad industry caution.
