How DeepMind Red‑Teams AI—and Why It Matters

About this episode

Dawn Bloxwich and Tom Lue from Google DeepMind discuss AI safety, red-teaming, and the regulatory landscape for frontier AI.

Two of Google DeepMind’s leaders—Dawn Bloxwich (Responsible Development & Innovation) and Tom Lue (VP, Frontier AI Global Affairs)—open the playbook on frontier safety and policy. Bloxwich explains how DeepMind blends structured evaluations with red‑teaming by experts and jailbreakers, and how its Frontier Safety Framework addresses severe risks (from CBRN and cyber to loss of control and socio‑affective concerns). Lue maps the regulatory landscape across the EU, US and Asia, arguing for harmonized standards that safeguard without stifling innovation. Together, they show what responsible development looks like before and after deployment—model cards, third‑party testing and the industry forums trying to align best practices. See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info .

People in this episode

Guests: Dawn Bloxwich, Tom Lue

Topics covered

• AI safety
• red-teaming
• regulatory landscape
• responsible development
• Frontier Safety Framework

Keywords

• DeepMind
• AI
• safety
• red-teaming
• regulation
• innovation
• Frontier Safety Framework

Mentioned in this episode

Organizations: Google DeepMind, EU, US, Asia

More episodes of The Tech Download

• Europe’s $14 Billion AI Challenger: Mistral CEO Arthur Mensch · June 3, 2026 · 46 min
• How AI Will Reshape Education—What Actually Works · January 22, 2026 · 33 min
• How DeepMind Plans to Reach AGI Safely · January 15, 2026 · 53 min
• The Tech Download: Inside Google DeepMind - Trailer · January 14, 2026 · 2 min
• Why cross-border payments are Asia’s next big tech frontier — Insights from an Indonesian unicorn · November 7, 2025 · 16 min
• VC founder: AI isn't a bubble — but its founders need to think globally · November 4, 2025 · 40 min