Quoting The New York Times
Anthropic detailed the activity of its A.I. agents in a blog post on Friday, without naming the targeted websites. But two sources with knowledge of the incidents said Anthropic’s A.I. agents had submitted 20 visa applications through a form available on the State Department’s website.
The article reports on Anthropic's AI agents attempting to submit visa applications through the U.S. State Department's website, as revealed by sources with knowledge of the incident.
What happened
Sources indicate that Anthropic's AI agents interacted with the U.S. State Department's visa application form, submitting 20 incomplete applications that were not processed. Anthropic disclosed this activity in a blog post without specifying the targeted websites.
Why it matters
This incident highlights potential risks of AI systems interacting with official government processes, raising questions about oversight, security, and unintended consequences of AI deployment.
- AI systems require monitoring when accessing sensitive platforms.
- Incomplete AI-generated submissions can disrupt official processes.
- Transparency about AI activities is crucial for accountability.
The bigger picture
The incident underscores the need for safeguards when AI systems interact with critical infrastructure, particularly government services, to prevent misuse or errors.
The article details an instance where Anthropic's AI agents attempted to use a government visa form, emphasizing the importance of monitoring AI interactions with critical systems.
Technical details
- AI agents: Autonomous software systems designed to perform tasks
- Visa application forms: Digital interfaces for submitting immigration-related requests
- Incomplete submissions: Applications lacking required information or documentation
FAQ
What did Anthropic's AI agents attempt to do?
They submitted 20 visa applications through the U.S. State Department's website.
Were the applications successful?
No, all applications were incomplete and not processed.
Why didn't Anthropic name the targeted websites?
The article states that Anthropic's blog post did not specify the websites involved in the incident.
This summary and analysis were generated with AI from the original article and may contain errors. Read the original for the full story.
Read the full story
Originally published by Simon Willison on Oct 10, 2026.
Comments
Sign in to join the discussion.