# Anthropic warns GLM-5.3 poses cyberattack risk

Digest AI · Policy & Regulation · published 2026-10-13T10:00:00Z

Canonical: https://digestai.news/story/anthropic-warns-glm-5-3-poses-cyberattack-risk

## Summary

Anthropic released a report late last month claiming that the open‑weight model GLM‑5.3, released in August by the Chinese company Z.ai, poses a major cyberattack risk. The report notes that the model is freely downloadable and modifiable, raising concerns about misuse.

The debate over open versus closed models is framed by contrasting views from industry leaders. Nvidia CEO Jensen Huang supports more open models, arguing that the U.S. should develop them to keep hardware sales high. Anthropic’s Dario Amodei cautions that cheap, powerful models could undermine the company’s position and pose safety risks.

Researchers and policy experts weigh the benefits of open models—greater accessibility, lower cost, and potential for wider security testing—against the dangers of unfiltered use and the possibility of catastrophic misuse. The discussion highlights the trade‑offs between democratization of AI and the need for robust safety measures.

## Key points

- Anthropic report claims GLM‑5.3 poses major cyberattack risk.
- GLM‑5.3 released by Z.ai in August and is freely downloadable.
- Nvidia CEO Jensen Huang supports more open models, while Anthropic’s Dario Amodei cautions against them.

## Why it matters

The open‑model debate shapes how AI power is distributed, affecting security, innovation, and the balance of influence between tech firms and governments.

## Sources

1. [Open models have real benefits. Do they outweigh the risks?](https://technologyreview.com/2026/10/13/1146366/open-models-have-real-benefits-do-they-outweigh-the-risks) (MIT Technology Review AI, 2026-10-13)

Part of the developing story: [Rogue AI Agents Breach Global Networks](https://digestai.news/thread/ai-agents-linked-to-openai-attempted-hacks-on-unm-data-usa-and-australian) (3 stories)

## Cite

Digest AI, "Anthropic warns GLM-5.3 poses cyberattack risk", 13 October 2026, https://digestai.news/story/anthropic-warns-glm-5-3-poses-cyberattack-risk

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/anthropic-warns-glm-5-3-poses-cyberattack-risk.json