|
Welcome back!
Major Anthropic customers are still waiting to get “zero data retention” for the newest Fable AI models, meaning the AI firm would not store data related to their usage of its models. In the meantime, some are turning to OpenAI.
Palantir and Booz Allen, for instance, are expanding their use of OpenAI’s Astra model, which they’ve been able to secure with zero data retention. Both have restricted their use of Fable models because Anthropic said in June it would store customer data for those using Fable models for 30 days, for security reasons.
Before that June announcement, Anthropic offered zero data retention policies in deals negotiated with individual customers. Starting later this fall, the AI firm plans to roll out a new program called Enterprise Frontier Safeguards, which effectively gives “eligible” customers zero data retention.
Exactly how a company qualifies for the program isn’t clear but Anthropic has identified some customers that helped create it including Salesforce, Comcast, and Visa.
(Anthropic’s website says it’s offering zero data retention guarantees for the Fable models to some of those customers while they wait for the program to go live.)
Anthropic’s policy towards customer data has lately become a lightning rod for criticism of the AI labs. Palantir CEO Alexander Karp, for instance, has blasted the AI firms, accusing them of using their customers’ sensitive corporate data to compete with them.
The issue is proving beneficial for OpenAI, which is trying to expand its enterprise business so it can better compete with Anthropic. Palantir began making Astra available to its customers earlier this month. That means many enterprises that use AI models through Palantir have been exposed to the capabilities of OpenAI’s Astra likely weeks before they’ll get to test out Anthropic’s Fable models. As we’ve seen recently, weeks in the world of AI development may as well be years.
An executive of one company that provides cybersecurity software to the US government and large financial services firms said it switched from using Anthropic’s models to run its customer-facing apps to primarily using models including Astra from OpenAI because it could get zero data retention guarantees from OpenAI that it couldn’t get from Anthropic.
To be sure, both Anthropic and OpenAI have been under pressure from customers and the public to disclose how their models learn from user data, particularly after OpenAI faced criticism over how its model may have learned from researchers’ work to achieve a mathematical breakthrough (OpenAI later denied doing so).
But any leg-up OpenAI can get is crucial at a time when Anthropic is dominating the enterprise AI space and both firms prepare to go public in the near future.
|
