Key Takeaways
- Anthropic’s Claude AI was withdrawn from Goldman Sachs Hong Kong banking personnel in recent weeks
- Goldman’s internal platform continues to provide access to alternative AI solutions like Gemini and ChatGPT
- Goldman determined Hong Kong staff should discontinue use of all Anthropic offerings following contract assessment
- Anthropic’s website does not include Hong Kong among its officially supported regions
- Financial institutions and regulators worldwide are examining potential risks associated with Anthropic’s latest Mythos model
Investment banking powerhouse Goldman Sachs has discontinued access to Anthropic’s Claude AI platform for its Hong Kong-based banking personnel, according to a source directly familiar with the matter.
The individual informed Reuters that Hong Kong-based staff members had previously utilized Claude via an internal artificial intelligence platform. Within recent weeks, this capability was terminated.
Alternative artificial intelligence platforms, including Google’s Gemini and OpenAI’s ChatGPT, remain accessible to personnel through the identical internal system.
The Financial Times initially disclosed the modification, referencing sources knowledgeable about the circumstances. Goldman Sachs chose not to provide commentary. Anthropic did not furnish an immediate response to comment requests.
Based on the FT’s reporting, Goldman arrived at this determination after conducting a rigorous interpretation of its contractual agreement with Anthropic. The financial institution engaged in discussions with Anthropic prior to determining that its Hong Kong workforce should cease utilizing any Anthropic offerings.
An Anthropic representative informed the FT that Claude had never received official “support” designation in Hong Kong. The organization’s official website excludes Hong Kong from the roster of markets where its API or Claude.ai receive official availability status.
Hong Kong’s Unique Position in AI Access
Hong Kong occupies a complex space regarding American AI platform accessibility. Mainland China prohibits AI solutions from U.S. corporations, while Hong Kong has generally maintained openness, with individual companies establishing their own access parameters.
This development surfaces amid escalating friction between the United States and China concerning AI technology and information security matters. A scheduled summit between Presidents Donald Trump and Xi Jinping in mid-May will reportedly address these subjects.
The United States government released a worldwide advisory in the preceding week regarding purported AI intellectual property theft by Chinese AI entities. During 2024, OpenAI implemented measures to block China-based traffic to its API amid concerns that Chinese rivals might exploit its models for training purposes.
Reuters was unable to verify whether additional financial institutions or corporations had implemented Claude access restrictions in Hong Kong.
Mythos Model Under Regulatory Examination
Goldman’s decision arrives as Anthropic’s most recent AI system, Mythos, has captured the attention of international financial institutions and regulatory authorities.
The Hong Kong Monetary Authority informed Reuters that it has initiated contact with multiple prominent financial institutions to gather information about recent Mythos developments. The authority has also instructed these organizations to refresh their risk evaluations and implement appropriate protective measures.
Reuters documented that certain lending institutions were already evaluating additional security protocols in response to cybersecurity concerns associated with sophisticated AI systems.
During February, Goldman Sachs Chief Information Officer Marco Argenti stated the institution was collaborating with Anthropic to create AI-driven agents designed to automate an expanding spectrum of operational functions.
The Hong Kong government provided no response to a Reuters inquiry seeking commentary. The HKMA refrained from commenting specifically on Goldman’s decision.
