Anthropic, AI Startup, Faces Data Breach and Federal Investigation

Anthropic, an AI startup known for its creation of the Claude family of large language models (LLMs) and associated chatbots, is experiencing a tumultuous week. On Monday, January 22nd, the company discovered that a contractor accidentally sent a file with non-sensitive customer information to a third party. The file contained a “subset” of customer names and open credit balances as of the end of 2023. An Anthropic spokesperson assured, “Our investigation shows this was an isolated incident caused by human error — not a breach of Anthropic systems. We have notified affected customers and provided them with the relevant guidance.”

Shortly after this incident, the Federal Trade Commission (FTC) announced an investigation into Anthropic’s strategic partnerships with tech giants Amazon and Google. The FTC is also examining the relationships between rival company OpenAI and Microsoft. However, Anthropic stated that the data breach is unrelated to the FTC probe and declined to comment further on the investigation.

“We sincerely regret that this incident occurred and any disruption it might have caused you,”

Anthropic spokesperson emphasized, “

Windows Report, a PC-centric news outlet, obtained and published an email from Anthropic to its customers acknowledging the leak of their information by a third-party contractor. In the email, Anthropic clarified that the leaked information included the customers’ account names and accounts receivable information as of December 31, 2023. The company emphasized that no sensitive personal data, such as banking or payment information, was leaked.

Anthropic assured its customers that the contractor’s actions were an isolated error and that there was no evidence of any malicious behavior resulting from the disclosure. However, the company urged customers to remain cautious and vigilant of any suspicious communications appearing to come from Anthropic, such as requests for payment or credentials. Customers were advised to ignore any suspicious contacts and follow their own internal accounting controls.

When asked about the impact of the data breach, an Anthropic spokesperson informed VentureBeat that only a “subset” of users were affected, without specifying the exact number. This breach is especially significant as data breaches caused by human error are currently at an all-time high, accounting for 95% of cases.

FTC Investigation and Concerns Over Data Security

Anthropic’s data breach reinforces concerns among enterprises that use third-party LLMs like Claude with their proprietary data. Recent events, including Samsung’s ban on ChatGPT due to leaked company data, have heightened worries among technical decision-makers in both large and small businesses.

Anthropic, a rival of OpenAI, achieved rapid success since its establishment in 2021. The company is currently valued at $18.4 billion and secured $750 million through three funding rounds last year. It is set to receive significant investments from Google ($2 billion) and Amazon ($4 billion). Additionally, Anthropic is engaged in discussions regarding another funding round led by renowned tech VP company Menlo Ventures.

However, the company’s relationships with AWS and Google have raised concerns with the FTC. The agency issued 6(b) orders to Anthropic, Amazon, Microsoft, OpenAI, and Alphabet, requesting detailed information about their multi-billion-dollar partnerships. Specific attention is given to the agreements, collaborations, and their implications on competition. The FTC aims to assess whether these ties grant dominant firms undue influence or privileged access, compromising fair competition.

The European Union and the UK are also investigating Microsoft’s AI investment, and the FTC expects information related to these probes as well. Anthropic has been an active partner with AWS and Google since its inception, engaging in extensive collaboration that has expanded over a short period of time.

Partnerships with Amazon and Google

Anthropic’s collaboration with Amazon includes a minority ownership by the e-commerce giant, which has invested up to $4 billion in the AI startup. AWS, Amazon’s cloud computing division, serves as Anthropic’s primary cloud provider and supplies chips to the startup. Anthropic has also committed to providing future generations of its models to AWS customers through Amazon Bedrock, offering early access to unique features for customization and fine-tuning purposes.

Regarding its partnership with Google and Alphabet, Anthropic uses Google Cloud security services, a PostgreSQL-compatible database, and the BigQuery data warehouse. The company relies on Google’s TPU v5e for its Claude large language model (LLM). Google Cloud CEO Thomas Kurian expressed confidence in the partnership, stating, “This expanded partnership with Anthropic, built on years of working together, will bring AI to more people safely and securely.”

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Posts