tech
OpenAI seeks to one-up Anthropic with new customer privacy protections
A competition is developing between OpenAI and Anthropic over who can provide the best privacy protections for enterprise customer data.
TL;DR
- OpenAI announced Private Safety Processing, a new service to detect AI misuse without retaining customer data.
- This system contrasts with Anthropic's policy of retaining user data for 30 days for safety analysis.
- OpenAI's approach expands on Zero Data Retention (ZDR) by monitoring multiple conversations for long-horizon misuse.
- Private Safety Processing uses an agent to analyze interactions across sessions for signs of abuse, even if spread out.
- If triggered, the system sends a signal to OpenAI, which may then contact the customer for more context.
- Anthropic allows human review of data through a controlled path with recorded sessions.
- The move is seen as part of a competitive rivalry between OpenAI and Anthropic.
- Recent reports indicate slower growth for OpenAI compared to Anthropic, fueling the competition.