Skip to main content
DutyStation
Search
🤖 AI & TechTechCrunchSaturday, October 10, 2026 at 12:18 AM UTC1h ago

Anthropic can’t reliably control its AI agents. It’s cutting off its internal evals from the live internet instead

Anthropic said it "turned off live internet access" for "all our internal evaluations" until further notice

Read full article at TechCrunch →

Originally published by TechCrunch. Summary and curation by DutyStation News.

📰 Related Stories

Free Gemini app users now only have access to ‘Auto’ models
🤖 AI & Tech

Free Gemini app users now only have access to ‘Auto’ models

9to5Google23m ago
Anthropic Agents Tried to Fill Out Visa Forms on State Dept. Website
🤖 AI & Tech

Anthropic Agents Tried to Fill Out Visa Forms on State Dept. Website

NYT Tech2h ago
🤖 AI & Tech

The maker of non-text AI model Jev valued at $7.5B just weeks after launch

TechCrunch4h ago
Anthropic’s AI gave Philadelphia police a fake tip about an unsolved homicide
🤖 AI & Tech

Anthropic’s AI gave Philadelphia police a fake tip about an unsolved homicide

The Verge4h ago
Powerful magnitude 7.7 earthquake hits Panama, damaging buildings
🤖 AI & Tech

Powerful magnitude 7.7 earthquake hits Panama, damaging buildings

BBC World6h ago
🤖 AI & Tech

An Anthropic AI model sent a false homicide tip to Philadelphia police

TechCrunch6h ago
← Back to News
Anthropic can’t reliably control its AI agents. It’s cutting off its internal evals from the live internet instead — DutyStation News