Anthropic claims popular Chinese AI model has Mythos-class hacking abilities
- ID
- 30340
- Status
- summarized
- Published
- 30 Sep 2026, 10:40 PM
- Fetched
- 30 Sep 2026, 11:05 PM
- Provider
- Tom's Hardware
- Category
- technology
- Original URL
- https://www.tomshardware.com/tech-industry/artificial-intelligence/anthropic-claims-popular-chinese-ai-model-has-mythos-class-hacking-abilities-frontier-red-teaming-report-details-weak-safeguards-on-open-weight-ai
- Source URL
- https://www.tomshardware.com/feeds/all
Summary
- Score
- 5.5
- Created
- 30 Sep 2026, 11:06 PM
- Tags
- Audience
- developersai_agent_usersstartup_founders
What happened
Anthropic published a report claiming Zhipu AI's GLM-5.3 can generate malicious content, be used for cyberattacks, and that its safeguards can be bypassed via "several methods." The Tom's Hardware news-analysis (by Sayem Ahmed, published 30 September 2026) frames this against Anthropic's own position as a closed-source lab eyeing an IPO whose CEO Dario Amodei has called for pacing the AI frontier, while Claude Opus 5.5 and Sonnet 5.5 shipped days after those alarms were raised. The excerpt names no specific bypass techniques, model version tested, or benchmark numbers.
Why it matters
If your agents or product route prompts through GLM-5.3 or other open-weight models, this is a competitor's claim published without methodology you can inspect in the text — so it is not grounds to swap providers. What it does change: expect enterprise buyers and procurement to ask which model version you pin and what guardrails sit in front of it, and plan your own eval of the exact checkpoint you deploy rather than relying on either lab's framing.
Discussion angle
What evidence would actually move you off a model you already ship with — named bypass techniques, a reproducible eval, a specific checkpoint hash? And does a closed-source lab's red-team report on an open-weight rival carry enough weight to act on, given the IPO and competitive context?