OpenAI announced in August 2026 a new privacy-focused safety system called Private Safety Processing, positioning it as a direct alternative to Anthropic’s recently introduced data-retention policy for enterprise customers.
Private Safety Processing is an automated, agent-driven system that monitors for potential misuse across multiple user sessions without retaining any customer data or requiring human review of conversations. OpenAI is currently previewing the service with select customers.
The announcement comes in response to Anthropic’s July 2026 policy requiring it to retain user session data — including full conversation histories — for 30 days on what the company calls “covered models,” which include all Mythos-class models such as Fable, and future models with similar capabilities. Anthropic said the policy exists for safety purposes, allowing it to analyze potential misuse. The policy has drawn concern from enterprise customers who handle sensitive data and do not want it stored or reviewed by the AI lab. Anthropic does permit human review of that data, but only through a controlled access path involving a small set of approved reviewers, with every review session recorded in a tamper-proof log.
Both companies already follow a Zero Data Retention (ZDR) approach for standard API usage, which monitors for abuse on a per-session basis without storing customer data. OpenAI describes Private Safety Processing as an extension of ZDR that enables long-horizon safety monitoring — assessing inputs and outputs across multiple conversations rather than a single session. This allows the system to detect bad actors who spread malicious requests, such as attempts to engineer malware, across multiple interactions to avoid detection.
If the system is triggered, it may send a narrowly defined signal to OpenAI indicating a specific type of activity. OpenAI can then determine whether enforcement is necessary and may contact the customer for additional context. Customers retain the choice of whether to share data with OpenAI at that point.
The competing privacy approaches reflect broader tension between the two companies. A recent report indicated OpenAI’s Q2 revenue growth was slower than Anthropic’s, whose annualized revenue run rate is now reportedly $65 billion. Both companies are also working toward IPOs, with Anthropic investors suggesting a potential valuation of $2 trillion.
Source: TechCrunch