There’s a reason 400,000 professionals read this daily.

Join 400,000+ professionals at Google, Microsoft, OpenAI, and more who start their day with The AI Report — the #1 B2B AI newsletter. We cut through the noise and show you exactly how to use AI tools to save time, boost productivity, and stay ahead at work.

No hype. No fluff. Just real-world use cases, expert prompts, and actionable tools.

👉 [Subscribe Now] — and start working smarter today.

Today:

• Google Launches Gemini 2.5 Pro Preview

• OpenAI To Cut Microsoft Revenue Share

• Amazon Developing AI Coding Tool Kiro

• xAI, Palantir Partner On Financial AI

• Google Launches AI Max For Ads

Google Launches Gemini 2.5 Pro Preview

Google's Gemini 2.5 Pro Preview (I/O edition) is now available with enhanced coding capabilities, particularly for building interactive web apps. Released ahead of schedule due to strong demand, it leads the WebDev Arena Leaderboard by +147 Elo points. The model excels at coding tasks like code transformation and complex workflows. It also features improved video understanding, scoring 84.8% on VideoMME. Developers can access it via Google AI Studio, Vertex AI, and the Gemini app.

Why It Matters

1. Enhanced Coding Efficiency: Gemini 2.5 Pro's improvements in coding and interactive web app building increase developer productivity, setting a new standard for AI-assisted coding.
   

2. Multimodal Capabilities: Its advanced video understanding and multimodal reasoning showcase the integration of diverse AI functions, vital for building interactive and dynamic applications.
   

3. AI Model Benchmarking: Leading the WebDev Arena Leaderboard highlights its superior performance, influencing industry standards for AI-driven web development tools.

OpenAI To Cut Microsoft Revenue Share

OpenAI plans to reduce the revenue it shares with Microsoft as part of a restructuring aimed at investor and regulatory satisfaction. The share, originally 20% through 2030, might drop to 10%. OpenAI seeks to give Microsoft traditional equity instead of profit rights while keeping access to AI technology beyond 2030. Tensions arise as Microsoft demands continued use of OpenAI’s IP and server integration while OpenAI negotiates partnerships with Oracle and SoftBank.

Why It Matters

1. Strategic Partnership Shift: Reducing revenue sharing could alter OpenAI’s long-term collaboration with Microsoft, impacting AI infrastructure and business models.
   

2. Financial Independence: OpenAI's move towards equity rather than revenue sharing reflects a desire for financial stability and reduced dependency on a single partner.
   

3. AI Ecosystem Impact: Changes in OpenAI’s revenue model may influence how tech giants structure partnerships with AI research firms, setting a precedent for future collaborations.

Amazon Developing AI Coding Tool Kiro

Amazon is developing an AI coding tool called Kiro, designed to streamline software development using AI agents and a multi-modal interface. Kiro can analyze prompts, generate code in real time, and integrate with both first-party and third-party AI agents. It aims to overcome limitations of existing coding tools by offering faster, more intuitive software creation. Kiro could significantly boost productivity and help Amazon compete in the rapidly growing AI coding assistant market.

Why It Matters

1. Revolutionizing Software Development: Kiro’s real-time code generation and multi-modal interface can drastically reduce development time, setting a new standard for AI-powered coding.
   

2. Competitive Edge: As tech giants like Microsoft and Google invest in AI coding tools, Kiro positions Amazon as a strong competitor in the AI developer ecosystem.
   

3. Agentic AI Integration: Kiro’s use of multiple AI agents for complex coding tasks reflects a shift towards more autonomous and adaptive AI development environments.

🧠RESEARCH

Voila: Voice-Language Foundation Models for Real-Time Autonomous Interaction and Voice Role-Play

Voila is a voice-language model designed for real-time, human-like interaction. It supports fast, expressive conversations with low latency (195 ms), blending language understanding with rich voice features like tone and emotion. Users can customize voices easily, and Voila also handles ASR, TTS, and multilingual translation, paving the way for dynamic, voice-based AI interactions.

RM-R1: Reward Modeling as Reasoning

RM-R1 introduces a new approach to reward modeling by treating it as a reasoning task rather than just scoring outputs. Using reasoning chains and reinforcement learning, it improves both interpretability and performance. RM-R1 outperforms larger models like Llama3.1-405B and GPT-4o by up to 13.8%, setting a new standard for reward models in LLM alignment.

Grokking in the Wild: Data Augmentation for Real-World Multi-Hop Reasoning with Transformers

Grokking in the Wild explores how to improve multi-step reasoning in transformers using data augmentation. By adding synthetic facts to existing knowledge graphs, even if some are incorrect, the model learns to reason rather than memorize. This method boosts accuracy on multi-hop reasoning tasks, achieving up to 100% on 2WikiMultiHopQA, surpassing current benchmarks.

🛠️TOP TOOLS

DeepSwapper - AI-powered face swapping tool that offers unlimited face swaps without watermarks or advertisements.

Kreado AI - Multilingual video creation, allowing users to generate content in over 140 languages and accents. 

Resemble.ai - Generative voice AI platform that enables users to create highly realistic synthetic voices for various applications.

BlueWillow - AI image generator that turns text prompts into visually stunning artwork or designs.

2Short AI - Creates short, engaging clips from longer videos, helping you quickly repurpose content for highlights or promotions.

📲SOCIAL MEDIA

🗞️MORE NEWS

• xAI and Palantir are partnering to help financial services use AI effectively. Their new "agentic workforce" includes AI agents for business processes. Despite high AI interest, banks face challenges with integration, including security concerns.

• Google launched AI Max for Search campaigns, enhancing targeting and creative capabilities with AI. The tool helps advertisers expand reach, optimize ads, and gain deeper insights. Early users like L'Oréal report doubled conversion rates and reduced costs.

• Hugging Face launched Open Computer Agent, a free, cloud-hosted AI tool similar to OpenAI’s Operator. While it handles basic tasks, it struggles with complex ones and CAPTCHAs. Its goal is to showcase affordable, capable open AI models.

• OpenAI is acquiring Windsurf, a popular LLM-powered developer platform, for $3 billion. Windsurf’s flexibility in supporting non-OpenAI models, like Meta’s Llama 3.1, raises questions about future integration. The acquisition aims to strengthen OpenAI’s position in AI-driven coding tools.

• Google introduced "Simplify" for its iOS app, using AI to make dense text easier to read without leaving the page. It leverages Gemini to retain key details while making complex information more understandable.

• Researchers at Penn Medicine used AI to speed up the discovery of antiviral drugs against human enterovirus 71 (EV71). The AI model identified effective compounds faster than traditional methods, offering a template for rapid antiviral development.

• Reddit will implement user verification to prevent AI bots from impersonating humans. The move follows an experiment where AI bots posted on a subreddit, sparking authenticity concerns. Reddit aims to balance verification with preserving user anonymity.