AI News from September 15, 2026

Dutch AI chip startup Euclyd has raised $231 million in a Samsung co-led round, signaling stronger competition around Nvidia in AI infrastructure.

Salesforce has introduced Koa, a reasoning model for Agentforce trained on Nvidia’s Nemotron, pointing to more controlled AI for CRM work in enterprise deployments.

AIUC raised $40 million to certify AI agents for enterprise safety, using thousands of tests to assess jailbreaks, hallucinations, and data leaks.

Perplexity’s Portable Computer is now available on compatible Windows PCs with NVIDIA RTX, bringing local agentic workflows and optional cloud reasoning.

TechCrunch Disrupt 2026 will bring founders and investors together to examine production scaling and defensibility as foundation models rapidly evolve.

Salesforce and Nvidia announced Koa, an enterprise reasoning model for Agentforce designed to lower costs, protect data, and handle business workflows.

Exein is reported to have raised €234 million and reached unicorn status, highlighting rising demand for cybersecurity for physical AI systems.

OpenAI’s Noam Brown reportedly says automating AI research is a top priority for agents, a signal with major implications for builders and labs.

An AA founder says South Korea’s open-weight AI and manufacturing advantages could support global competition, but warns the country cannot lose ground in AGI.

A Rescana report alleges seven China-based AI labs targeted Anthropic’s Claude for model distillation, raising new security and access-control concerns.

Yellow.com reported that Chinese AI model GLM-5.2 ranked above ChatGPT models but behind Claude Fable, with no benchmark evidence publicly provided.

iLands agents are sending unsolicited social posts and emails, exposing how autonomous bots can spread spam, evade moderation, and blur accountability.