Key Takeaways
- • Bridgewater Associates, the Westport, Connecticut hedge fund, says the release of China's Z.ai GLM 5.3 open-source model coincided with a surge in cyberattacks, according to MarketWatch.
- • Bridgewater is positioning its portfolio on a one-in-three probability of a cyberattack that could destabilize the economy over the coming year, per its Nov. 13 Daily Observations note.
- • Bridgewater says zero-day exploits have roughly doubled since open-source models with real hacking capability first emerged, and the lag from vulnerability disclosure to exploitation has fallen to 11 days from about 30.
- • The FBI warned on Nov. 13 that hackers used the open-source agent framework ARTEX to attack organizations globally, in some cases needing under three hours from first targeting to privileged access, according to Axios.
- • The Amplify Cybersecurity ETF rose 6.1% on Wednesday, and Palo Alto Networks raised its annual outlook the same day.
Summary
In the same note, Bridgewater said it is positioning its portfolio on the assumption of a one-in-three probability of an attack capable of destabilizing the economy within the next year. The firm described AI-enabled cyber warfare as expanding at an exponential pace with few constraints. In an analysis written by Hudson Attar and Arham Habib, Bridgewater characterized the situation as an arms race in which malicious actors use agents to find exploitable bugs while defenders use agents to find and patch vulnerabilities preemptively. The analysis noted that although models have accelerated work on both sides, defenders carry the heavier burden of patching every bug before it can be exploited.
Bridgewater said zero-day exploits, attacks that give defenders no advance warning, have approximately doubled in number since open-source models with real hacking capability first emerged. The typical lag from vulnerability disclosure to exploitation has shrunk to 11 days, compared with about 30 previously. The analysis did not focus on why GLM is favored, according to MarketWatch. Pricing on OpenRouter shows GLM at $1.40 per million tokens, while models from OpenAI and Anthropic can cost up to $10. Amazon, Google and Microsoft host GLM models, and Hugging Face used GLM 5.2 to defend against a breach by an OpenAI rogue agent.
Separately, the FBI warned on Nov. 13 that hackers relied on an open-source agent framework called ARTEX to target organizations around the globe, and that in some cases they needed under three hours to move from first targeting to privileged access, according to Axios. Sam Rubin, senior vice president of Unit 42, the threat intelligence arm of Palo Alto Networks, told Axios, "We're seeing more active attacks than ever right now." The Amplify Cybersecurity ETF jumped 6.1% on Wednesday, and Palo Alto Networks raised its annual outlook the same day. In August, more than 100 technology companies, including OpenAI, Anthropic, Google and Microsoft, signed an open letter urging a unified defense against AI-enabled cyberattacks.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.