Microsoft launches AI testing initiative with OpenAI and Anthropic

Key Takeaways

Trump Admin Signs AI Deals With Microsoft Google and xAI

The Trump Administration has finalized new agreements with Microsoft, Google DeepMind, and xAI to test their AI models before public release. The Center for AI Standards and Innovation will partner with these companies to evaluate national security risks and safety. Developers are sharing models with reduced safeguards to help government evaluators understand potential threats. This collaboration aims to build trust in advanced AI systems while keeping pace with rapid technological advancements.

US Government Will Safety Test New AI Models From Tech Giants

Google, Microsoft, and xAI have agreed to let the US Department of Commerce test their new AI tools before they reach the public. This move expands on previous deals made during the Biden Administration and includes companies like OpenAI and Anthropic. The evaluations will cover testing, research, and safety best practices for commercial AI systems. Officials say this partnership is critical for understanding the security implications of frontier AI technology.

Tech Giants Allow Government Testing of Unreleased AI Models

Google, Microsoft, and xAI will share unreleased versions of their AI models with the government to improve cybersecurity. The National Institute of Standards and Technology announced this initiative to ensure AI systems are secure and trustworthy before public launch. The partnership includes developing new testing methods and a common framework for evaluating AI safety. This step is expected to significantly improve the reliability of AI systems used in critical infrastructure.

US Gains Pre-Release Access to AI Models From Five Major Labs

The US government now has access to test AI models from five major technology labs before they are released to the public. The Center for AI Standards and Innovation signed new agreements with Google DeepMind, Microsoft, and xAI for national security testing. Officials state that rigorous measurement science is essential for understanding the risks of advanced AI. These collaborations help scale the center's work in the public interest during a critical time for AI development.

Big Tech AI Investments Reach Nearly $1 Trillion Mark

Big Tech companies have invested nearly $1 trillion in artificial intelligence, with returns showing in improved efficiency and productivity. A report by CB Insights found the top 10 investors have poured $950 billion into AI since 2010. Despite these successes, AI data centers face challenges from local opposition and power grid shortages. In South Korea, the power grid is struggling to meet the growing demand for electricity needed to support the industry.

AI Spending Surges Past $830 Billion Amid Grid Challenges

Big Tech companies have invested $830 billion in AI research, intensifying competition but creating power grid constraints. Google leads with $20 billion in investments, followed by Microsoft and Amazon. Local communities are opposing data center construction due to noise and environmental concerns. Experts warn that governments must invest in grid development to support the growing demand for AI electricity.

Startup Illoca Raises $13 Million to Bring AI to Architecture

Illoca, a startup using AI to help architects design buildings, has raised $13 million in new funding. Investors including Khosla Ventures and Founders Fund led the round to help the company expand its team and develop new products. The company plans to use the money to grow its operations and enhance its AI tools for the architecture industry.

Meta Tests Employee Keystrokes to Train AI Models

Meta is capturing employee keystrokes and mouse movements to train its AI models on how humans navigate computers. The Model Capability Initiative will collect this data on company devices without an opt-out option. Meta states the data will only be used for training and not for performance evaluations. Experts warn this practice raises privacy concerns, especially since the US lacks a federal privacy act.

VA Uses AI to Speed Up Veterans Disability Claims Processing

The Department of Veterans Affairs is using AI to process disability claims faster than ever before. A veteran named Chip used ChatGPT to help write his claim and received a 100% rating within a month and a half. The VA has cut its claims backlog by more than 70% since last year through automation and hiring. However, some former employees worry that speed does not always guarantee accuracy in complex disability decisions.

Nvidia and Corning Partnership Could Shift AI Infrastructure

Jim Cramer highlights a partnership between Nvidia and Corning as a potential seismic shift for AI infrastructure. Corning plans to increase optical fiber manufacturing capacity tenfold to replace copper inside data centers. This change could reduce heat constraints in Nvidia's next-generation systems. Both company CEOs are scheduled to appear on Mad Money to discuss the deal.

Class of 2026 Faces Tough Job Market and AI Worries

The Class of 2026 enters a job market with a four-year high unemployment rate for recent graduates. Nearly 9 in 10 graduates worry that artificial intelligence will replace entry-level roles. Many students are turning to trades like welding or nursing for more stable careers. Despite these fears, nearly 40% of graduates are using AI tools to help them find jobs.

AMD Q1 Earnings Show Datacenter Sales Driving AI Growth

AMD reported strong Q1 2026 results with datacenter sales contributing over half of its revenue. Major deals with Meta and partnerships with Google, Microsoft, and Tencent are fueling this growth. The company projects a 28% increase in earnings per share compared to the previous year. However, supply chain risks and high stock prices create potential profit-taking opportunities.

FBI Tracks AI Scams as Financial Losses Reach Record Highs

The FBI has started tracking AI-related fraud for the first time, revealing billions in losses. Americans reported nearly $21 billion in cybercrime losses in 2025, with over $893 million tied to AI scams. Scammers use AI to clone voices and create realistic messages to trick victims. Texas ranked second in the nation for both the number of complaints and total financial losses.

ChatGPT Now Integrated Into Excel and Google Sheets

ChatGPT has officially integrated with Microsoft Excel and Google Sheets to help users analyze data. This feature allows users to transform raw data into leadership insights using natural language prompts. The integration is available globally for all subscription plans and connects to external data sources like S&P Global. This move makes advanced data analysis more accessible to users without technical expertise.

Sources

• Trump Administration Finalizes AI Agreements with Microsoft, Google DeepMind, and xAI
• US to safety test new AI models from Google, Microsoft, xAI
• Microsoft, Google and xAI will let the government test their AI models before launch
• US government now has pre-release access to AI models from five major labs for national security testing
• Big Tech AI Investments Near $1 Trillion as Returns Materialize
• Big Tech AI Investments Surge Past $830 Billion
• Exclusive: Illoca raises $13M to bring AI to architecture
• Meta's AI training with keystrokes: Progress or privacy issue
• Hiring, Overtime, and AI: VA Is Processing Veterans’ Disability Claims Faster Than Ever
• Jim Cramer: This partnership could be a 'seismic shift' in AI and 2 of our stocks
• Class of 2026 faces tough job market and AI concerns as graduation season approaches
• AMD Q1 2026 Earnings: Datacenter Sales Mark Shift In AI Purchase Cycles
• The FBI Is Now Tracking AI Scams, and the Losses Are Huge
• ChatGPT Integrates with Excel and Google Sheets