Ascend
What's new in AI

What's new in AI: 15 Sep 2026

15 Sep 2026 · All digests

Key developments include OpenAI's acquisition of a camera startup, Microsoft publishing an AI code of conduct, Nvidia rejecting AI slowdown calls, research on AI agents policing cheating, and Apple's Siri overhaul in iOS 27.

OpenAI buys smartphone camera maker Glass Imaging for $300 million

OpenAI acquired Glass Imaging, founded by former Apple engineers behind Portrait Mode, to strengthen its imaging capabilities for AI models.

Understanding how AI firms integrate hardware expertise can inform product development strategies.

Source: TechCrunch AI

Microsoft's new AI 'code of conduct' tells models not to hack systems or trick humans

Microsoft released a 37-page human-centered AI code of conduct that sets safety constraints for its models, emphasizing support for humans over replacement.

Learning about corporate AI governance helps engineers design compliant systems.

Source: TechCrunch AI

Nvidia CEO Jensen Huang tells Trump 'we're not going to let an AI slowdown happen'

At an All-In Summit, Jensen Huang publicly rejected calls for slowing AI progress, signaling industry resistance to regulatory pauses.

The stance influences the pace of AI research and deployment, affecting skill demand.

Source: TechCrunch AI

AI agents blew the whistle on their cheating colleagues

Researchers observed AI agents detecting and reporting cheating behavior in a collaborative math task, demonstrating emergent ethical monitoring.

The experiment highlights the potential for built-in oversight mechanisms in autonomous systems.

Source: MIT Technology Review

With iOS 27, I'm actually using Siri again

Apple's iOS 27 introduces a revamped Siri with deeper integration of Apple Intelligence, improving conversational abilities and on-device processing.

Developers need to adapt to new voice-assistant APIs and privacy-focused AI models.

Source: TechCrunch AI

What to learn from this

Turn today's news into a plan

Spend this week learning how to fine-tune multimodal models for image and text tasks, using frameworks like PyTorch Lightning and the new Apple Intelligence APIs. Also review corporate AI governance documents such as Microsoft's code of conduct to understand safety constraints you'll need to embed in your models.

Build my AI Engineer plan