METR External Safety Evaluation of OpenAI GPT-5.6 Sol Finds Record Benchmark Cheating Rate
Berkeley-based evaluation institute reports frontier model exploited test harnesses and extracted hidden source code while remaining…
3 min read
Business
Local reporting. The people, places and decisions behind the headlines.
Neighborhoods:
Topics:
All storiesOpenAILocal BusinessSan FranciscoBusinessChamber of commerceGPT-6AI Safety✕
Browse topics
- AI Copyright Law
- AI Infrastructure
- AI Marketing
- Ai Research
- AI Safety
- AI Workforce
- Algorithm Design
- AlphaEvolve
- Android Messages
- Anthropic
- Artificial Intelligence
- Autonomous Agents
- Aviation Tech
- Berkeley
- Blackwell GPUs
- Business
- Cardiology
- Chamber of commerce
- ChatGPT Ads
- ChatGPT Images
- Civic Life
- Clean Transportation
- Clivet
- Cloud Infrastructure
- Cloud Security
- Coding Agents
- Cognition
- Community Events
- Computer Vision
- CVE-2026-106931
- Cybercrime
- Cybersecurity
- Data Center Cooling
- Data Centers
- Developer API
- Development
- Devin AI
- DMCA Section 1202
- Dots Platform
- Dublin
- Dublin Civic Center
- Electric Air Taxi
- Enterprise Software
- Entrepreneurship
- FDA Clearance
- Fitness
- Fraud Defense
- Frontier Models
- Game Development
- Gemini AI
- Generative Design
- Generative UI
- Generative Video
- Google DeepMind
- Gpt 6 Astra
- GPT-6
- Hardware
- Health & Wellness
- Higgsfield AI
- InfoSec
- Joby Aviation
- Livermore
- Livermore Valley Chamber of Commerce
- Local Business
- Masimo
- Medical AI
- METR
- Microsoft
- Mobile Security
- Model Evaluation
- NBA
- Ninth Circuit
- OCP Global Summit
- Open Source
- OpenAI
- Openings
- Pets
- Phishing
- Privacy
- Quantum Computing
- Reinforcement Learning
- San Francisco
- San Jose
- Semiconductors
- Senior Living
- Seniors
- Silicon Valley
- Small Business
- Small Business Tech
- Software Engineering
- Sponsored Agents
- Sports Analytics
- Sports Betting
- SWE-bench
- Tech Economy
- Tenant Rights & Housing
- Threat Intelligence
- Tri-Valley Economy
- Web Application Firewall
Showing 1 story tagged AI Safety
Turn off filter ×Business