Microsoft’s Cybersecurity AI Push, Nadella’s Multi-Model Warning, and Medical Multimodal Benchmarks | UpNext AI – July 28, 2026

Today on UpNext AI, Microsoft makes a bigger play in AI security, Satya Nadella argues companies should not trust a single AI provider for everything, and a new medical AI paper tests whether multimodal systems can handle clinical work that depends on both images and text.

Covered stories:
- Microsoft launches its first cybersecurity-specialized model and a new agentic security platform
- Satya Nadella warns companies against relying on one AI model provider for all of their needs
- New ClinFusion paper evaluates a vision-centric medical multimodal model across a broad benchmark suite
- Verizon touts a reported $1 billion dark-fiber deal tied to Google data centers
- A new report says Hugging Face-hosted image models can be used to create nonconsensual deepfakes

Source links:
- Microsoft launches its first cybersecurity model, plus a new agentic cybersecurity system — https://techcrunch.com/2026/07/27/microsoft-launches-its-first-cyber-model-and-a-new-agentic-cybersecurity-system/
- Satya Nadella says companies that trust one AI for everything may not survive — https://techcrunch.com/2026/07/27/satya-nadella-says-companies-that-trust-one-ai-for-everything-may-not-survive/
- ClinFusion: A Vision-Centric Multimodal LLM System for Holistic Medical Understanding — https://arxiv.org/abs/2607.24743v1
- Verizon touts $1B dark fiber deal for Google data centers as first of many — https://arstechnica.com/ai/2026/07/verizon-seeks-ai-profits-with-mini-data-centers-1b-dark-fiber-deal-with-google/
- Hugging Face is being used to easily undress women and children — https://www.theverge.com/ai-artificial-intelligence/971723/hugging-face-nudify-deepfake-undress-women-children