Google's AI-Powered Search Infrastructure Faces Scrutiny

Google's AI-powered search infrastructure is facing scrutiny for giving the company an unfair advantage over other AI firms. A British regulator and industry players are pushing back against Google's dominance, citing concerns that its strategy is killing the web. This has raised questions about the fairness of Google's practices.

Meanwhile, Michigan has formed the Michigan Artificial Intelligence Association, a new AI trade group aimed at positioning the state as a national leader in AI. The group will work with businesses, academia, and government to advance AI research and development.

In the realm of AI security, frontier AI models are generating a large number of vulnerabilities, making it difficult for security teams to prioritize and fix them. To address this, AWS' Rudra Mitra suggests applying AI power to prioritize, fix, and validate vulnerabilities.

AI-powered security threats are also increasing, and managed service providers (MSPs) need to rethink their customer protection strategies. AI is being used to create more convincing phishing attacks and to expose sensitive business data.

Siemens has set limits on its physics AI technology, which can explore thousands of design variations quickly but cannot be used for safety-critical applications. The technology is designed to work with human oversight.

On the education front, an AI-powered private school is expanding across the country and is considering Kansas City as its next location. The school uses AI to provide personalized and effective learning experiences for students.

In other developments, Sakana AI has tested its Gemma 4 model for orchestration, and Tencent has expanded international access to its Hy3 AI large language model. The market for LLM observability and evaluation platforms has also grown significantly.

Finally, AI-powered technology is helping people who are blind or have low vision to achieve their goals, from reading menus to running a half marathon.

Key Takeaways

• Google's AI-powered search infrastructure faces scrutiny for giving the company an unfair advantage over other AI firms. • Michigan forms the Michigan Artificial Intelligence Association to promote AI development and adoption. • Frontier AI models generate a large number of vulnerabilities, making it difficult for security teams to prioritize and fix them. • AI-powered security threats are increasing, and MSPs need to rethink their customer protection strategies. • Siemens sets limits on its physics AI technology for safety-critical applications. • An AI-powered private school is expanding across the country, including potentially Kansas City. • Sakana AI tests its Gemma 4 model for orchestration. • Tencent expands international access to its Hy3 AI large language model. • The market for LLM observability and evaluation platforms has grown significantly. • AI-powered technology is helping people who are blind or have low vision to achieve their goals.

Google's AI Strategy Sparks Web Fightback

Google's AI-powered search infrastructure is giving the company an unfair advantage over other AI firms. This has led to concerns that Google's strategy is killing the web. A British regulator and industry players are now pushing back against Google's dominance. The company's strategy involves using its search infrastructure to strengthen its AI products. This has raised questions about the fairness of Google's practices.

Google's AI Dominance Sparks Concerns

Michigan Forms AI Trade Group

A new AI trade group, the Michigan Artificial Intelligence Association, aims to position Michigan as a national leader in AI. The group is headed by William Hallan and Thomas Clement of the Michigan Retailers Association. The goal is to promote AI development and adoption in Michigan. The group will work with businesses, academia, and government to advance AI research and development.

Frontier AI Requires Smarter Vulnerability Triage

Frontier AI models are generating a large number of vulnerabilities, making it difficult for security teams to prioritize and fix them. To address this, AWS' Rudra Mitra suggests applying AI power to prioritize, fix, and validate vulnerabilities. This involves using AI to analyze and prioritize vulnerabilities based on their severity and potential impact.

AI Security Risks Push MSPs to Rethink Customer Protection

AI-powered security threats are increasing, and managed service providers (MSPs) need to rethink their customer protection strategies. AI is being used to create more convincing phishing attacks and to expose sensitive business data. MSPs need to help customers govern AI use, strengthen trust and access controls, and prepare for AI-enabled attacks.

Siemens Sets Limits on Physics AI

Siemens' physics AI technology can explore thousands of design variations quickly, but it cannot be used for safety-critical applications. The technology is designed to work with human oversight, and Siemens is clear about its limitations. The company's goal is to provide engineers with a tool that can help them make design predictions quickly and accurately.

AI-Powered Private School Eyes Kansas City

An AI-powered private school is expanding across the country and is considering Kansas City as its next location. The school uses AI to provide personalized and effective learning experiences for students. The founder of the school believes that AI can help address concerns about education and provide better outcomes for students.

Sakana AI Tests Gemma 4 for Orchestration

Sakana AI has tested its Gemma 4 model for orchestration and found it to be effective. The model is part of Sakana's Fugu platform, which provides a unified platform for multi-agent orchestration. The test demonstrates the flexibility of Sakana's model pool and conductor architecture.

Tencent Expands Access to Hy3 AI Model

Tencent has expanded international access to its Hy3 AI large language model. The model is now available to developers, businesses, and individual users through several Tencent products and third-party platforms. This expansion aims to increase the adoption of AI technology globally.

Top LLM Observability and Evaluation Platforms

The market for LLM observability and evaluation platforms has grown significantly in 2026. These platforms provide tools for monitoring and evaluating AI models, including tracing, evals, and production monitoring. The top platforms include Langfuse, LangSmith, Braintrust, Arize, and more.

AI-Powered Technology Helps Visually Impaired People

AI-powered technology is helping people who are blind or have low vision to achieve their goals. From reading menus to running a half marathon, AI is providing new opportunities for visually impaired individuals. The technology is being developed to improve the lives of people with visual impairments.