2026-07-21

AI/Tech Digest 2026-07-21

Safety and alignment in an era of long-horizon models [LAB]
OpenAI details new safety risks and safeguards emerging from the deployment of autonomous, long-running AI agents.
https://openai.com/index/safety-alignment-long-horizon-models (openai.com)
Introducing Cosmos 3 Edge [LAB]
Hugging Face announces a new model release focused on edge computing efficiency.
https://huggingface.co/blog/nvidia/cosmos3edge (huggingface.co)
Trump administration weighs blocking US access to Chinese AI models ... [web]
Potential US regulatory action against Chinese AI models could significantly shift the competitive landscape for domestic players like OpenAI and Anthropic.
https://www.hindustantimes.com/world-news/us-news/trump-administration-weighs-blocking-us-access-to-chinese-ai-models-what-it-means-for-openai-anthropic-101784565266821.html (hindustantimes.com)