Startup Uses AI Agents To Access Anthropic Data Within Hours
(MENAFN) A cybersecurity startup reportedly accessed sensitive internal information belonging to artificial intelligence company Anthropic in less than 12 hours after identifying weaknesses in third-party business software used by the company.
Cathedral, a recently established startup founded by former members of the US Department of Government Efficiency (DOGE), discovered vulnerabilities that provided potential entry points into systems used by the developer of the Claude chatbot, according to reports.
A draft report reviewed by media outlets said the startup deployed AI agents to identify routes into Anthropic’s internal environment. The researchers reportedly gained access to meeting transcripts, salary information, job offers and Slack conversations concerning advanced AI models.
The work was carried out through a program that permits external cybersecurity researchers to test Anthropic’s systems and identify security vulnerabilities.
Anthropic confirmed that Cathedral had accessed a “limited set” of its sales, human resources and information technology systems.
The company said it had secured the exposed access routes and informed federal officials about the incident.
An Anthropic spokesperson said Cathedral had spent several days trying to find a way into the company’s own software before shifting its attention to third-party business tools.
The incident has reportedly drawn attention within the US government. A senior Trump administration official said the report had circulated among government officials and raised concerns about how quickly the startup had reached sensitive systems.
The official also warned that the findings raised questions about whether foreign intelligence agencies could take advantage of comparable security weaknesses.
Cathedral, a recently established startup founded by former members of the US Department of Government Efficiency (DOGE), discovered vulnerabilities that provided potential entry points into systems used by the developer of the Claude chatbot, according to reports.
A draft report reviewed by media outlets said the startup deployed AI agents to identify routes into Anthropic’s internal environment. The researchers reportedly gained access to meeting transcripts, salary information, job offers and Slack conversations concerning advanced AI models.
The work was carried out through a program that permits external cybersecurity researchers to test Anthropic’s systems and identify security vulnerabilities.
Anthropic confirmed that Cathedral had accessed a “limited set” of its sales, human resources and information technology systems.
The company said it had secured the exposed access routes and informed federal officials about the incident.
An Anthropic spokesperson said Cathedral had spent several days trying to find a way into the company’s own software before shifting its attention to third-party business tools.
The incident has reportedly drawn attention within the US government. A senior Trump administration official said the report had circulated among government officials and raised concerns about how quickly the startup had reached sensitive systems.
The official also warned that the findings raised questions about whether foreign intelligence agencies could take advantage of comparable security weaknesses.
Legal Disclaimer:
MENAFN provides the
information “as is” without warranty of any kind. We do not accept
any responsibility or liability for the accuracy, content, images,
videos, licenses, completeness, legality, or reliability of the information
contained in this article. If you have any complaints or copyright
issues related to this article, kindly contact the provider above.
Comments
No comment