👀 Microsoft swaps OpenAI for Anthropic?

Wednesday’s AI Report

• 1. 👀 Microsoft swaps OpenAI for Anthropic?
• 2. 🏠 Invest in this real estate start-up, Pacaso
• 3. 🌍 How ESPN expanded sports coverage with AI
• 4. 💼 Build an AI consultancy career with Innovating with AI
• 5. 💡 Partner Perspectives
• 6. ⚙️ Trending AI tools
• 7. 🧠 NVIDIA reveals biggest AI chip yet
• 8. ⚖️ Judge halts Anthropic’s $1.5B settlement
• 9. 📑 Recommended resources

Read Time: 5 minutes

❓Wednesday’s Partner Perspective: Are NVIDIA’s scaling opportunities within NVIDIA itself, or in its supply chain? Scroll down to find out what The AI Report’s Director of Partnerships, Kyle Mair, thinks ⬇️

✅ Refer your friends and unlock rewards. Scroll to the bottom to find out more!

Microsoft swaps OpenAI for Anthropic?

🚨 Our Report

Microsoft has struck a partnership with one of OpenAI’s biggest rivals, Anthropic, to use its AI models to power features within its Office 365 suite. This move highlights Microsoft's intent to reduce reliance on one single-provider, OpenAI, which could further intensify potential tensions between the two.

🔓 Key Points

Anthropic’s latest AI model, Claude Sonnet 4, will power certain features in Microsoft’s Word, Excel, Outlook, and PowerPoint applications, which were previously, solely, powered by OpenAI’s GPT-5 model.
Insiders revealed that this isn’t a bargaining tactic from Microsoft: Rather, Microsoft feels that Anthropic’s models outperform OpenAI’s in areas like automating tasks in Excel and generating effective PowerPoints.
Microsoft already uses Anthropc’s Claude and xAI’s Grok to power GitHub Copilot, and it recently released its first, in-house models (MAI-Voice-1 and MAI-1-preview), to diversify and reduce reliance on providers.

🔐 Relevance

Reports about rising tensions between OpenAI and its biggest backer, Microsoft, for many months, as both make moves to reduce dependency and become more independent: OpenAI, for instance, is reportedly building its own infrastructure to reduce reliance on Microsoft’s Azure Cloud, has plans to mass-produce its own AI chips with Broadcom, and is even venturing into areas that directly compete with Microsoft, including building a LinkedIn alternative (LinkedIn is a subsidiary of Microsoft). Despite this, Microsoft insists that “OpenAI will continue to be our partner on frontier models and we remain committed to our long-term partnership."

FULL STORY

Keep this stock on your watchlist

They’re a private company, but Pacaso just reserved the Nasdaq ticker “$PCSO.”

Created by a former Zillow exec who sold his first venture for $120M, Pacaso brings co-ownership to a $1.3T real estate market.

They’ve generated $1B+ worth of luxury home transactions & fees across 2,000+ owners. That’s good for more than $110M in less than 5 years.

No surprise the same firms that backed Uber and Venmo already invested in Pacaso.

But you don’t have to be a Wall Street firm to invest. Pacaso is offering everyday investors the same prospect, and 10,000+ people have already joined them.

And you can join them for just $2.90/share.

But don’t wait too long. Invest in Pacaso before September 18.

INVEST HERE

Read the offering circular at invest.pacaso.com

NVIDIA’s Supply Chain - Part 1

By Kyle Mair · Director of Partnership at The AI Report
The AI Report Partner Perspectives Column

READ KYLES’ PERSPECTIVE

How ESPN expanded sports coverage with AI

ESPN, a US sports media brand, wanted to scale its coverage by including under-served sports, without overtaxing editorial resources.
They deployed an AI system that analyzed structured data like box scores, play-by-play logs, rosters, schedules, and audio transcripts.
The system produced game recaps for under-covered sports, so ESPN just needed to review AI copy for tone and accuracy, before publishing.
Recaps are now regularly generated for overlooked sports, broadening content reach, without incurring huge increases in editorial effort.

FIND OUT MORE

Want to build a 6-figure business as an AI consultant?

The AI consulting market is about to grow by a factor of 8X – from $6.9B to $54.7B in 2032.

But how do you turn your AI enthusiasm into marketable skills, clear services and a proper business?

Our friends at Innovating with AI have trained 1,000+ AI consultants – and their exclusive consulting directory has driven Fortune 500 leads to graduates.

Enrollment in The AI Consultancy Project is opening shortly – and you’ll only hear about it if you apply for access.

GET EARLY ACCESS

Omega turns hours of prospect research into seconds with a Custom AI Sales Assistant ⭐️⭐️⭐️⭐️ (G2)
MagicVest is an AI-powered crypto safety & signal tool
AIEnhancer transforms blurry photos into stunning visuals with AI

At the annual AI Infrastructure Summit, NVIDIA, revealed its newest chip, called Rubin CPX, which has been specifically designed for context windows bigger than 1M tokens, and will be available next year.
What this means is that AI models powered by this new chip will be able to handle and process around 750 words at once (which is the equivalent of about 3,000–4,000 book pages), before giving its response.
For context, older models like GPT-3.5, for example, could process just 12,000 words before forgetting earlier parts, meaning this new chip could perform better on tasks like video generation or software development.

FULL STORY

Earlier this week, we reported that Anthropic had agreed to pay $1.5B to three authors, as part of a lawsuit which alleged the start-up had breached copyright laws by illegally training their AI models on pirated books.
Now, federal Judge, William Alsup, needs “further information,” before he can approve the settlement, and warned both parties that if he doesn’t receive this information by September 25, the case will proceed to court.
Judge Alsop slammed the proposed settlement as “nowhere near close to complete,” claiming he felt “misled” by a lack of detail, particularly around whether the settlement funds would actually reach the authors.

FULL STORY