Anthropic co-founder and CEO Dario Amodei speaks throughout a man-made intelligence panel at Inbound 2025 Powered by HubSpot, on the Moscone Center in San Francisco, Sept. 4, 2025.
Chance Yeh | Getty Images Entertainment | Getty Images
Anthropic on Tuesday introduced it should change a controversial data retention policy for its enterprise customers after “a lot of feedback,” and it mentioned it plans to implement a brand new resolution known as Enterprise Frontier Safeguards as a substitute.
The firm launched its present data retention policy in June alongside the launch of Claude Fable 5 and Mythos 5, two of the corporate’s most superior artificial intelligence fashions. Anthropic mentioned on the time that it might require 30-day retention for all site visitors on these fashions to assist defend towards misuse and “complex and novel” cyberattacks.
Anthropic pledged to not use that data for any “non-safety-related purpose,” together with mannequin coaching, however lots of its enterprise customers expressed concern.
Kate Jensen, head of Americas at Anthropic, advised CNBC that the corporate spent a whole bunch of hours working with customers to develop one other resolution.
With the launch of Enterprise Frontier Safeguards, Anthropic mentioned companies will be capable to management how their data is reviewed, saved and managed, and they’ll additionally be capable to perform automated security monitoring the place no Anthropic human assessment is required. The firm mentioned it won’t cost for Enterprise Frontier Safeguards, and that the controls will work whether or not customers entry Anthropic’s expertise instantly or by means of a cloud supplier.
“We can do the scanning that we really need to, to help everybody feel confident, but it’s on their systems and their data so that they don’t have to compromise on their own privacy posture,” Jensen mentioned in an interview.
Anthropic’s June data retention policy nonetheless applies to its non-enterprise subscribers that use its Mythos-class fashions.
Founded in 2021, Anthropic generates the overwhelming majority of its income by promoting to enterprises, and the market is essential because it gears up for what’s extensively anticipated to be a massive IPO. The firm can be racing to say forward of rivals like OpenAI, which launched its personal data retention solution earlier this month.
Last week, Anthropic CEO Dario Amodei previewed Enterprise Frontier Safeguards throughout an interview with CNBC concerning the firm’s expanded partnership with Salesforce. Amodei mentioned the answer helps make sure that Salesforce’s buyer data stays non-public, and that Anthropic’s fashions “don’t run out of control.” He added that Anthropic put a “huge amount of effort” into managing permissions for the combination between the 2 firms.
Jensen mentioned she feels “very confident” that Anthropic will see accelerated deployment following the announcement, particularly on customers’ most delicate workloads. Anthropic additionally unveiled its two newest fashions on Tuesday, Claude Fable 5.1 and Mythos 5.1, which the corporate mentioned present improved coding, data work and analysis capabilities.
Enterprise Frontier Safeguards will roll out in phases, and Anthropic mentioned it is aiming for broader availability this fall.
WATCH: Anthropic pushes into physical world with new standard to help AI agents operate machines
