Anthropic Plans to Change Data Retention Policy for Advanced AI

Anthropic PBC plans to allow business customers to keep greater control of their data when using its most capable artificial intelligence models, a stark change from an earlier data retention policy intended to mitigate potential cyberattacks.The Claude chatbot maker expects to roll out a new safety system later this year that will still require enterprise customers to retain data for 30 days but with the option to do so on their own cloud computing infrastructure rather than Anthropic’s, according to a person familiar with the matter. The system has been in the works for months, said the person, who spoke on condition of anonymity as the information is not public.Anthropic declined to comment.Anthropic announced in June that it would retain all customer data for its more powerful Mythos and Fable models — as well as as future frontier offerings — for 30 days to help spot and prevent new kinds of cyberattacks using its technology. The company said at the time that it wouldn’t use the data to train its AI. Still, Anthropic acknowledged in a recent report that the policy would “be unpopular with customers who have come to expect zero retention, and pose real risks to our business success (especially if competitors do not follow).”Longtime Anthropic rival OpenAI said on Wednesday that it began testing its own method, called private safety processing, to avoid retaining customer data from users. The company is trying out the software with companies including Databricks Inc. and major OpenAI investor Microsoft Corp., and plans to release it more widely in September.Anthropic began working on the newer system before users began expressing criticism, said the person. It has been coordinating with more than 100 customers in highly regulated industries to develop the option, the person added.