Inherent says its AI agent Faraday beat OpenAI and Anthropic at paper replication, using a small model and reinforcement learning.
OpenAI now wants California’s AI safety bill strengthened, calling for more monitoring and cybersecurity in frontier model development.
New research shows frontier AI labs still lack public rogue model containment plans as regulators push for stronger disclosure.
Trending
Why Inner Mongolia Has Become China’s AI Data Center Capital
Latest
Weekly Top
DOJ’s A16z Probe Puts Venture Board Seats Under the Microscope
AI Agents
173News
AI Fields
40News
AI Startups
320News
Amazon
27News
Anthropic
251News
Apple
58News
Applications
512News
Artificial Intelligence
1033News
Chatbots
158News
Companies
883News
Compute
191News
Computer Vision
6News
Deep Learning
11News
DeepSeek
4News
Energy
73News
Entertainment
62News
Ethics & Society
421News
Funding
163News
Gaming
5News
140News
Healthcare
20News
Industries
152News
Intel
3News
Large Language Models
555News
Legislation & Government
280News
Machine Learning
47News
Manufacturing
19News
Meta
44News
Microsoft
156News
Natural Language Processing
14News
NVIDIA
126News
Open Source
157News
OpenAI
481News
Originals
78News
Privacy & Security
341News
Reinforcement Learning
5News
Research
280News
Retail
16News
Robotics
46News
Safe Superintelligence Inc
4News
Search
103News
Sports
4News
Stability AI
3News
Supercomputing
64News
SuperIntelligence
33News
Tesla
12News
Transport
19News
Uncategorized
4News
Virtual Assistants
122News
Voice Recognition
4News
X.ai
55News
- 08 / 22 : DeepMind Alumni Startup Inherent Says Small AI Agent Beat OpenAI and Anthropic at Research Replication
- 08 / 22 : OpenAI Urges California to Tighten Its AI Safety Law After Earlier Opposition
- 08 / 22 : AI labs face pressure over rogue-model containment as regulators demand answers
- 08 / 21 : Why Inner Mongolia Has Become China’s AI Data Center Capital
- 08 / 21 : Anthropic’s Claude Opus 4.6 Fails Sexual-Content Safeguards in Repeated Tests




