cyb3rops: open-weight AI models mean attackers can compromise 90% of orgs without new exploits
cyb3rops · x · 2026-10-27
Security researcher cyb3rops (Florian Roth) clarifies his widely quoted remark about open-weight AI models: combining and spreading them into countless attack tools has turned attacks that once required advanced skills into something anyone can execute — but his point was not about exploit development.
His core claim: you don't need to discover a new software vulnerability to compromise 90% of organizations. He expects the quote to appear in industry slide decks in the coming months. The defensive takeaway: the real risk isn't 0days, it's open models dramatically lowering the barrier for automated social engineering and routine attacks.
More from Models
- OpenAI verifies cryptosystems safe from latest math result, trend alarming — IgorCarron · 2026-10-27
- DeepSeek Flash impresses dev: let it build its own tools, cheap enough to experiment — victormustar · 2026-10-27
- Reddit asks: Why do almost all major open-weight models come from China? — PhilosopherSweaty826 · 2026-10-27
- Doctor's Prescription OCR Test: Gemma 4 31B Now Reads What Models Couldn't a Year Ago — MaziyarPanahi · 2026-10-27
- Study Finds GPT-6 Astra Only Partially Capable at CAD Workflows — ZhejiangUniversity · 2026-10-27
- Anthropic investigates elevated error rates for Claude Sonnet 5.5 — ClaudeAI-mod-bot · 2026-10-27