Nando de Freitas: trusting models trained by others can be catastrophic, openness is essential for safe AI
DimitrisPapail · x · 2026-11-11
Nando de Freitas (VP, Google DeepMind) amplifies a study on backdooring LLM agents, arguing it is catastrophic to trust models trained by someone else and that openness and scrutiny are essential for safe AI. Security researcher John Langford adds that people should maintain custody of a model from creation, just as with binaries — otherwise models can be set up to exhibit arbitrary behaviors in nearly undetectable ways.
Related event: Experts Warn of Catastrophic LLM Backdoor Risks(2 posts)→
More from Safety
- Researcher Warns Real-Time Video AI Companions Are the Next Teen-Safety Wave — anshulkundaje · 2026-11-11
- 35,000+ Grok-built apps exposed patient data, plaintext passwords, and sales pipeline — The Decoder · 2026-11-11
- Fortinet patches nearly 100 vulnerabilities after adding AI models to security testing — Potential-Couple-745 · 2026-11-11
- New BadEdit backdoors LLMs via model editing with just 15 samples, up to 100% success — terryyuezhuo · 2026-11-11
- Treat weights like code: undetectable malicious payloads in open models spark safety debate — DimitrisPapail · 2026-11-11
- Alice & Bob: ~25,000 cat qubits could factor RSA-2048 in 3 days, slashing prior estimates — AryHHAry · 2026-11-11