AI news · 10 stories

AI news for Nov 19, 2026

OpenAI releases GPT-6.2 Astra and Sol · Google rolls out Genie 4 to AI Ultra subscribers · Microsoft launches MAI Code 2 coding models

News card: OpenAI releases GPT-6.2 Astra and Sol
OpenAI

OpenAI releases GPT-6.2 Astra and Sol

OpenAI introduced GPT-6.2 Astra and Sol for ChatGPT Work, Codex, and its API. The company reports better professional-task performance and token efficiency, with lower Astra API prices than GPT-6 Astra.

Read at OpenAI
News card: Google rolls out Genie 4 to AI Ultra subscribers
Google

Google rolls out Genie 4 to AI Ultra subscribers

Google began rolling out Genie 4 in Project Genie to eligible AI Ultra subscribers on November 18. Real-time environment editing, native audio, and a community gallery expand how subscribers create and share interactive worlds.

Read at Google
News card: Microsoft launches MAI Code 2 coding models
Microsoft

Microsoft launches MAI Code 2 coding models

Microsoft announced Code 2 and Code 2 Flash at Ignite on November 17, with general availability in GitHub Copilot and Foundry. The models target complex repository work and faster routine edits respectively, giving developers additional choices for coding-agent workloads.

Read at Microsoft
News card: Palo Alto Networks agrees to acquire Applied Compute
Palo Alto Networks

Palo Alto Networks agrees to acquire Applied Compute

Palo Alto Networks announced a definitive agreement on November 18 to acquire Applied Compute, a platform for post-training and inference of open-weight models. The planned combination aims to help enterprises adapt and deploy AI securely across public and sovereign clouds.

Read at Palo Alto Networks
News card: Isomorphic Labs raises $5 billion for AI drug design
PR Newswire

Isomorphic Labs raises $5 billion for AI drug design

Isomorphic Labs announced a $5 billion Series C on November 17, co-led by Greenoaks and Sequoia Capital. The funding will support its drug-design engine, medicine-candidate pipeline, and AI-led bioengineering infrastructure.

Read at PR Newswire
News card: NVIDIA details prompt-driven Cosmos training-data workflows
NVIDIA

NVIDIA details prompt-driven Cosmos training-data workflows

NVIDIA described Cosmos Data Factory on November 18 with five reference blueprints and six composable agent skills. The workflows coordinate generation, augmentation, and labeling of physical-AI training data, including driving scenarios and visual-inspection defects.

Read at NVIDIA
News card: Google shares lessons from Sashiko AI code review
Google Cloud

Google shares lessons from Sashiko AI code review

Google published findings on November 18 from Sashiko, its open-source Gemini-powered code-review system for Linux. Its benchmark-driven development and review experience highlight ways AI can help maintainers catch bugs amid growing volumes of generated code.

Read at Google Cloud
News card: Microsoft and Intel expand Azure infrastructure partnership
Microsoft

Microsoft and Intel expand Azure infrastructure partnership

Microsoft and Intel announced a long-term expansion of their Azure infrastructure collaboration, detailed in a November 18 post. The agreement includes Intel EMIB packaging for Azure first-party silicon and multi-year Xeon supply capacity to support cloud and AI demand.

Read at Microsoft
News card: Microsoft launches cross-industry AI Transparency Initiative
Microsoft

Microsoft launches cross-industry AI Transparency Initiative

Microsoft announced an AI Transparency Initiative with more than 40 organizations on November 17. The collaboration aims to develop open approaches to independently verifying and governing model and agent behavior across providers, with future integration into Microsoft products.

Read at Microsoft
News card: OpenAI rolls out Live Dictation in ChatGPT
OpenAI

OpenAI rolls out Live Dictation in ChatGPT

OpenAI announced Live Dictation on November 18, turning speech into text while users speak and removing the ten-minute recording limit. The feature is rolling out globally across all ChatGPT plans on mobile, web, and desktop, reducing waits for transcription.

Read at OpenAI

Get the next episode first

Shorts and deep dives on AI system design land on @AI.JoinDev first. Subscribe, or follow the RSS feed.