Week 36, 2026 AI roundup

Models H Company Releases NeoMME: A Family of 260M and 800M Single-Tower Multimodal Encoders That Drop the Vision Tower and Causal Decoder MarkTechPost Agents Authors push back as publishers and agents make claims on Anthropic settlement TechCrunch AI Models The Sequence Radar - Issue 927: Last Week in AI: Model Madness: The Frontier Has a Refresh Button The Sequence Models An Alien Mind OpenAI News Agents Research acceleration: The view inside OpenAI OpenAI News Models Seattle Times and Newsday are the latest publications to sue OpenAI and Microsoft TechCrunch AI Agents OpenAI agents discussed ways to escape their sandbox on public wiki Ars Technica AI Models Architecting memory and storage in the AI era MIT Technology Review AI Security Once popular for attacking AI, ASCII smuggling is embraced by spammers Ars Technica AI Security OpenAI Touts GPT-6 Astra as Its Safest Model, But It's Still Dangerous AI Business Models Anthropic’s $2 trillion IPO puts powerful external trustees in spotlight Ars Technica AI Policy Prompt: Nvidia Moves up the AI Stack AI Business