Radio
Now Playing
Quickyla Radio — Click to play
Open →
3 min left
Back to News

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

Anthropic said its models exploited websites on the internet, including some run by U.S. government agencies, and it will turn off live internet access for all of its internal evaluations until the f…

Anthropic can’t reliably control its AI agents. It’s cutting off its internal evals from the live internet instead
TechCrunch — 9 October 2026
Text:
6 0 0

Anthropic said its models exploited websites on the internet, including some run by U.S. government agencies, and it will turn off live internet access for all of its internal evaluations until the frontier lab is sure it can monitor and control its AI agents.

The incidents, disclosed in a blog post , involved AI agents tasked to solve problems seeking resources on the internet. In the process, they exploited software flaws, accessed databases without paying fees, used URL shortening services to smuggle information past restrictions, and even submitted a false murder tip to the Philadelphia police.

Anthropic said it discovered these new issues in a review of its model’s activities that began in July, demonstrating the lab’s lack of awareness of its software’s behavior in real time.

Notably, the company said that alignment training was not yet sufficient for skills like search and computer use that are central to its pitch that AI agents will be used by any professional who relies on digital tools.

The behaviors Anthropic disclosed are similar to incidents involving OpenAI agents that collaborated to break into various websites in search of information, including some run by the Australian government.

Anthropic previously disclosed that its models had broken into external systems. The frontier lab said it considered today’s disclosures “significantly less severe from an alignment and security perspective” than those it announced before.

However, the lab still said it had “turned off live internet access” for “all our internal evaluations” until it is certain it can monitor and control its agents.

It’s not clear what that means. Sydney Von Arx, the founder of Nightingale, an AI safety organization, told TechCrunch in an interview before this disclosure that developing models on a data center cut off from the open internet would be very challenging for researchers, and hinder the progress of the models, which benefit from internet access.

Read Full Story at TechCrunch →
Advertisement
React:
Sources
Sponsored

More to Read

When are surround sound systems actually worth it over soun…
💻 Technology
When are surround sound systems actually worth it over soundbars?
Engadget · 13 days ago
How to turn your old MacBook into a home NAS server
💻 Technology
How to turn your old MacBook into a home NAS server
Engadget · 13 days ago
China introduces strict IPO criteria for humanoid robot sta…
💻 Technology
China introduces strict IPO criteria for humanoid robot startups
CNBC Finance · 12 days ago
China posts weakest industrial profit growth this year, exp…
📈 Markets & Finance
China posts weakest industrial profit growth this year, expanding 4.2% in August
CNBC Economy · 13 days ago
South Korea’s exports hit record high on AI boom
🌍 World News
South Korea’s exports hit record high on AI boom
Al Jazeera · 10 days ago
UNGA81: Why has Africa’s Security Council reform push remai…
🌍 World News
UNGA81: Why has Africa’s Security Council reform push remained unresolved?
Al Jazeera · 15 days ago
Full view