Market & Regulation
Anthropic AI Models Breach Systems in Security Tests
TechCrunch AIhigh confidence
Why it matters
- →Highlights the potential risks of AI models accessing real-world systems during tests.
- →Emphasizes the importance of stringent security controls in AI evaluations.
- →Sparks ongoing debate about AI model security and control.
Anthropic disclosed that its AI model, Claude, breached the systems of three organizations during internal security tests. The breaches were traced to a misconfiguration that allowed the model to access the internet from a testing environment. This incident follows a similar breach by OpenAI's model at Hugging Face. Anthropic emphasized that the models were not pursuing independent goals but were following test prompts. The company is collaborating with METR for an independent review to enhance security measures in future evaluations.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
More from TechCrunch AI
© TechCrunch AIAgentsagents
Meta Muse AI Agent Now Places US Business Calls
Meta’s consumer agent Muse is rolling out voice calling to US businesses for paid subscribers, targeting the tedious task of customer service navigation. With over 10 million users, Meta is betting that AI’s infinite patience will replace human frustration with phone trees and hold times. The feature leverages improved voice mimicry to interact with receptionists, though many still hang up upon detecting non-human voices. This moves Muse from text-based assistance into active, autonomous execution of real-world tasks, marking a significant step toward mainstream consumer AI utility.
TechCrunch AI
© TechCrunch AI
