Story cards stay below the sticky dock while audio, chapters, date, and brief navigation remain accessible.
OpenAI is shutting down its Sora AI video generator and redirecting the freed compute toward robotics research. The move signals a sharp strategic pivot: video generation is being deprioritized as the company doubles down on physical AI systems.
Source: [crescendo.ai](https://www.crescendo.ai/news/latest-ai-news-and-updates)
---
Shopify merchants can now sell directly inside ChatGPT, Google AI Mode, Microsoft Copilot, and the Gemini app via "Agentic Storefronts." This is the most concrete commercial deployment of the Agentic Commerce Protocol to date, collapsing the gap between product discovery and purchase inside AI assistants.
Source: [crescendo.ai](https://www.crescendo.ai/news/latest-ai-news-and-updates)
---
Anthropic's Claude can now take over your cursor — pointing, clicking, scrolling, and navigating your desktop when no direct integration exists. Paired with "Dispatch," a mobile feature that lets users assign tasks from their phone while Claude executes them on a desktop, it's the most aggressive push yet into autonomous computer use.
Source: [crescendo.ai](https://www.crescendo.ai/news/latest-ai-news-and-updates)
---
OpenAI has surpassed $25 billion in annualized revenue and is reportedly taking early steps toward a public listing, potentially as soon as late 2026. Anthropic is not far behind at ~$19B ARR. The gap between these two and every other AI lab is widening fast.
Source: [llm-stats.com](https://llm-stats.com/ai-news)
---
The White House published its regulatory framework for AI, emphasizing innovation and light-touch oversight while threatening to withhold $21B in broadband funding from states that enforce AI laws deemed incompatible with federal policy. The federal-vs-state regulatory war over AI is now fully engaged.
Source: [Nextgov/FCW](https://www.nextgov.com/artificial-intelligence/2026/03/white-house-releases-regulatory-vision-ai/412274/)
---
Google's TurboQuant announcement allows large AI models to run in a fraction of their previous memory footprint, prompting a slight dip in DDR5 RAM prices. If the technique generalizes, it could dramatically reduce inference costs and broaden hardware accessibility for on-device AI.
Source: [crescendo.ai](https://www.crescendo.ai/news/latest-ai-news-and-updates)
---
OpenAI opened a formal safety bug bounty program inviting researchers to identify AI abuse and safety risks across its products. The move comes as the company faces increasing scrutiny over dual-use risks from its increasingly capable models.
Source: [crescendo.ai](https://www.crescendo.ai/news/latest-ai-news-and-updates)
---
**Big Picture**
Today's news is a story about AI growing up in two directions at once: outward into commerce and physical environments (Shopify storefronts, Claude's mouse control, robotics compute), and inward toward accountability (bug bounties, federal regulatory frameworks, bias audit mandates). The infrastructure layer is quietly tightening too — Google's TurboQuant compressing memory costs while OpenAI eyes a public market. The race is no longer just about benchmark scores; it's about who can monetize, govern, and physically embed AI into the world first.
---
Sources:
- [Latest AI News 2026 — crescendo.ai](https://www.crescendo.ai/news/latest-ai-news-and-updates)
- [LLM News Today March 2026 — llm-stats.com](https://llm-stats.com/ai-news)
- [White House AI Regulatory Vision — Nextgov/FCW](https://www.nextgov.com/artificial-intelligence/2026/03/white-house-releases-regulatory-vision-ai/412274/)
- [12 AI Models in One Week — arturmarkus.com](https://www.arturmarkus.com/12-ai-models-launched-in-one-week-during-march-10-16-2026-openai-google-xai-anthropic-mistral-and-cursor-compress-developer-selection-cycles-to-monthly-as-frontier-model-releases-pile-up/)
- [AI Regulations 2026 — Credo AI](https://www.credo.ai/blog/latest-ai-regulations-update-what-enterprises-need-to-know)
