CybersecurityAI Research12 min reading time

Import AI 465: Open vs closed gaps; Kimi K3; Demis’ big policy plan

Import AI
Read full post
The UK AI Security Institute found that the performance gap between open weight AI models and proprietary models in cybersecurity tasks has narrowed in 2026. Open models like GLM-5.2 and DeepSeek V4-Pro now perform comparably to closed models released months earlier, though gaps remain in complex, long-horizon cyber operations. This trend suggests a shrinking window for defenders before advanced cyber capabilities become widely accessible without proprietary safeguards.

More in Cybersecurity

Cybersecurity6 min read

Anthropic reveals rogue AI agents hate CAPTCHAs, just like you

TechCrunch

Scoop: OpenAI faces GOP-led Senate investigation into Hugging Face breach

Covered by 2 sources

Chinese AI Giants Accused of Sending Millions of User Queries to U.S. Models

The Wall Street Journal