🧠 OpenAI ignites new AI race with GPT‐5 and $38B Amazon deal

Welcome back to the AI Business Summary newsletter!

The Big Lead: OpenAI secures a $38 billion cloud services agreement with Amazon Web Services.

Here’s everything else you need to know this week in AI...

In Today’s Issue:

💰 OpenAI signs $38 billion AWS infrastructure deal

🚀 OpenAI launches GPT-5 with built-in thinking

🎨 Microsoft debuts first in-house image generator, MAI-Image-1

🗺️ Google Maps integrates Gemini as "all-knowing copilot"

⚖️ Amazon sues Perplexity over "agentic" shopping tool

🇨🇳China bans foreign AI chips from state-funded data centers

🏥 FDA advisers weigh therapy chatbot regulation

⚖️ Getty wins UK lawsuit against Stability AI over copyright violations

Frontier Model Power Plays

🚀 OpenAI launches GPT-5 with built-in thinking

• Significant leap in intelligence across coding, math, writing, and health

• A unified system that decides when to think longer

• 94.6% on AIME 2025, 74.9% on SWE-bench Verified

• 45% fewer hallucinations than GPT-4o; 80% fewer than o3

💰 OpenAI signs $38 billion AWS infrastructure deal

• Multi-year partnership for advanced AI workloads starting immediately

• Hundreds of thousands of NVIDIA GPUs, with scaling capability

• Custom Amazon EC2 UltraServers with GB200s and GB300s

• All capacity targeted for deployment by the end of 2026

🎨 Microsoft debuts first in-house image generator, MAI-Image-1

• Available in Bing Image Creator and Copilot Audio Expressions

• Excels at photorealistic food, nature, and lighting effects

• Joins MAI-Voice-1 and MAI-1-preview in Microsoft's model lineup

• Not yet available in the EU, "coming soon" per AI chief

Big Tech AI Integrations

🗺️ Google Maps integrates Gemini as "all-knowing copilot"

• Conversational route planning with open-ended questions

• Uses landmarks and businesses instead of distance directions

• Proactive traffic alerts for routine commutes

• Google says grounded in real datasets to prevent hallucinations

🛒 Amazon sues Perplexity over "agentic" shopping tool

• Amazon claims Comet AI covertly accesses customer accounts

• Accuses startup of disguising automated activity as human browsing

• Perplexity calls it "bullying" to block innovation

• Highlights the emerging debate over AI agent regulation

Global AI Geopolitics

🇨🇳 China bans foreign AI chips from state-funded data centers

• New guidance requires domestically-made chips for state projects

• Projects under 30% complete must remove foreign chips

• Affects Nvidia H20, B200, and H200 processors

• Boosts market share for Huawei, Cambricon, and other locals

AI Regulation & Governance

🏥 FDA advisers weigh therapy chatbot regulation

• Digital Health Advisory Committee meeting on generative AI risks

• Focus on "therapy devices" built on large language models

• Considering prescription vs over-the-counter contexts

• Part of a broader effort to clarify AI medical device rules

⚖️ Getty wins UK lawsuit against Stability AI over copyright violations

• Landmark ruling confirms training on copyrighted images without consent is infringement

• Court orders Stability AI to disclose model training data and pay damages

• Raises global stakes for generative AI copyright compliance

• Could set a precedent for the US and EU regulatory frameworks

🌱 Bonus Thought

88% of organizations use AI regularly, yet only one-third have scaled beyond pilots. Companies experiment with agents and chase efficiency gains, but struggle to redesign workflows for enterprise impact. Meanwhile, geopolitics fractures supply chains as China bans foreign chips and Amazon battles Perplexity over agentic commerce.

This signals maturation, not failure. High performers aren't just deploying more AI tools. They're transforming business processes, setting innovation goals beyond cost-cutting, and hiring AI talent while expecting workforce reductions. They're also preparing for risks they're finally experiencing firsthand.

The pattern emerges across every breakthrough: Speed versus safety. Innovation versus regulation. Competition versus collaboration. OpenAI scales thinking models while regulators debate therapy chatbots. Microsoft builds proprietary image generators while startups face lawsuits over automated shopping.

Redesign workflows now, or watch competitors capture the enterprise value you're piloting.

📝 Business Prompt to Try

Why It Works

• Strategic foresight: Prepares you for AI infrastructure shocks driven by geopolitics or export controls.

• Cost-conscious: Limits restructuring within a defined 10% budget variance, keeping the plan realistic.

• Operational resilience: Prioritizes continuity of compute access and scalability under shifting regulations.

• AI-tested: Uses simulation prompts to validate infrastructure and vendor strategy before making high-cost moves.

💡 Quote of the Week