THURSDAY · 19 NOV 2026 · 06:51 UTCRSSMY BRIEFINGS
POLICY20 AUG 2026 · 09:26 UTC · 2 MIN READ

How Does Zero Data Retention Work in AI Models?

OpenAI's Zero Data Retention policy ensures that eligible API customers' prompts and model responses are not retained after a request is processed, and a new feature called Private Safety Processing strengthens safeguards without compromising data privacy. This approach allows customers to maintain control over their data while still benefiting from advanced AI safety features.

POLICY DESK
How Does Zero Data Retention Work in AI Models?
OpenAI's Zero Data Retention policy ensures that eligible API customers' prompts and model responses are not retained after a requ…
How Does Zero Data Retention Work in AI Models?

OpenAI is sticking to its guns with Zero Data Retention for eligible API customers - a policy that ensures customer prompts and model responses are ditched after a request is processed. No lingering customer content for OpenAI personnel to review, and no sneaky model training using enterprise customer data unless they explicitly give the green light. This is huge for organizations dealing with sensitive info, as it lets them keep a tight grip on their data while still tapping into advanced AI features.

The introduction of Private Safety Processing is a game-changer for AI safety. It lets automated systems spot patterns across related interactions without giving OpenAI personnel a peek at the underlying content. This feature tackles the limitations of existing safety systems, which only look at interactions one by one and might miss potential misuse that only becomes clear when you look at multiple interactions together. Private Safety Processing works with customer content wherever it's stored - whether that's on the customer's own infrastructure or in OpenAI's storage - and it's all encrypted using keys controlled by the customer.

As AI models take on more complex tasks, the need for safety systems to evolve becomes glaringly obvious. When models handle longer, more intricate tasks, serious risks might only become visible after multiple interactions. Private Safety Processing is designed to mitigate these risks while keeping Zero Data Retention intact, which gives customers peace of mind knowing their data is protected and not being used without their consent. By building stronger safeguards and keeping customer info under their control, OpenAI is showing it's committed to making sure artificial general intelligence benefits everyone.

Source: OpenAI

Reson8 Releases Rumble-2-Flash 0.5B Speech Model for 40 Languages
models

Reson8 Releases Rumble-2-Flash 0.5B Speech Model for 40 Languages

Reson8 released Rumble-2-Flash, a 0.5-billion-parameter speech recognition model covering 40+ languages. Launched November 18, 2026 on Hugging Face under Apache 2.0, it runs in standard Transformers and vLLM without custom code. Distilled from the larger Rumble-2 via the GOLD on-policy method, it scores 6.96% error on the multilingual Open ASR Leaderboard, beating Meta's 7B omniASR-LLM and Whisper large-v3, and processes an hour of audio in about half a second on a single GPU.

Osmo Claims Its AI Model Matches Human Smell Ability
models

Osmo Claims Its AI Model Matches Human Smell Ability

Osmo, a Google Brain spinout, claims its AI model achieves "human-level olfactory intelligence," detecting and discriminating smells like an average person and creating fragrances comparable to a junior perfumer. Founder Alex Wiltschko says the tech, trained on over ten million human scent labels, is already used by major fragrance firms and identified repellents stronger than DEET. Critics argue marketing claims lack documentation and the system struggles with contextual smell interpr

Google Launches Free Tasks AI Agent Replacing Spark
models

Google Launches Free Tasks AI Agent Replacing Spark

Google has rebranded its Spark AI agent as Tasks and made it free for users in the US, Canada, Japan, Korea, Australia, and India. Powered by Gemini 4 Hydrogen, the agent handles planning chores like meal plans and calendar organization. US users gain additional capabilities, including calling businesses, making purchases via Google Pay, and delegating tasks via email or SMS. Each session runs in an isolated virtual machine, pausing for confirmation before sensitive actions.

Share this post
Report