Meta locks down news training data and centralizes AI delivery as OpenAI eyes a GitHub rival
Meta is formalizing AI training data access and centralizing AI deployment while OpenAI reportedly builds a GitHub rival, signaling a consolidation of data and developer infrastructure in the AI era. Meta signed a multimillion‑dollar licensing deal with News Corp to use its journalism for AI training, a move that reduces legal risk and improves content provenance for models like Llama while setting a template for paid data pipelines [Meta–News Corp deal](https://www.webpronews.com/meta-writes-a-big-check-to-rupert-murdoch-inside-the-multimillion-dollar-ai-deal-that-could-reshape-news-licensing/). In parallel, Meta is forming an Applied AI Engineering team to move research into production across Facebook, Instagram, WhatsApp, and hardware, centralizing deployment and accelerating model-to-feature delivery at consumer scale [Applied AI Engineering team](https://www.webpronews.com/meta-quietly-builds-a-new-ai-army-inside-the-creation-of-the-applied-engineering-team-that-could-reshape-how-billions-use-technology/). Meanwhile, OpenAI is reportedly developing a full coding platform to compete with GitHub, challenging Microsoft’s developer stronghold and hinting at AI‑native source control and collaboration workflows that could reshape CI/CD integrations and code review practices [OpenAI vs. GitHub](https://www.webpronews.com/openais-quiet-push-into-developer-tools-puts-it-on-a-collision-course-with-microsofts-github/).