AI, Data & Tech ... Simplified!

Weekly News Summary

Week of July 21, 2025 - AI News Roundup

The latest developments in artificial intelligence

Articles from July 21 - July 27, 2025

📋 Quick View

📰 Full Stories

Altman warns there’s no legal confidentiality when using ChatGPT as a therapist

📰 TechCrunch AI | 📅 July 25, 2025

In response to a question about how AI works with today's legal system, Altman said one of the problems of not yet having a legal or policy framework for AI is that there's no legal confidentiality for users' conversations.

Anthropic deploys AI agents to audit models for safety

📰 AI News | 📅 July 25, 2025

Anthropic has built an army of autonomous AI agents with a singular mission: to audit powerful models like Claude to improve safety. As these complex systems rapidly advance, the job of making sure they are safe and don’t harbour hidden dangers has become a herculean task.

New AI architecture 100x faster than LLMs

📰 Venture Beat | 📅 July 25, 2025

Singapore-based AI startup Sapient Intelligence has developed a new AI architecture that can match, and in some cases vastly outperform, large language models (LLMs) on complex reasoning tasks, all while being significantly smaller and more data-efficient. The architecture, known as the Hierarchical Reasoning Model (HRM), is inspired by how the human brain utilizes distinct systems for slow, deliberate planning and fast, intuitive computation.

White House released "Winning the Race: America's AI Action Plan"

📰 US Government | 📅 July 23, 2025

The U.S. aims to secure global AI dominance through rapid innovation, massive infrastructure development, and strategic international leadership, positioning AI as a driver of economic growth, national security, and technological breakthroughs. Its AI Action Plan emphasises deregulation, workforce benefits, unbiased systems, and protection against misuse to ensure America sets global AI standards.

Alibaba’s new Qwen model makes open-source records

📰 Alibaba on X | 📅 July 23, 2025

Alibaba released Qwen3-Coder-480B-A35B-Instruct, 480B-parameter Mixture-of-Experts model (35B active) natively supports 256K context and scales to 1M context with extrapolation. It achieves top-tier performance across multiple agentic coding benchmarks among open models, including SWE-bench-Verified.

OpenAI and Oracle announce Stargate AI data centre deal

📰 OpenAI News | 📅 July 22, 2025

Oracle and OpenAI have entered an agreement to develop 4.5 gigawatts of additional Stargate data center capacity in the U.S. This investment will create new jobs, accelerate America’s reindustrialization, and help advance U.S. AI leadership.

Google’s newest Gemini 2.5 model aims for ‘intelligence per dollar’

📰 AI News | 📅 July 22, 2025

Google just dropped the stable version of Gemini 2.5 Flash-Lite and they’ve essentially created a model that’s designed to be the workhorse for developers who need to build things at scale without breaking the bank. Building cool things with AI can often feel like a frustrating balancing act.

OpenAI and UK Government announce strategic partnership to deliver AI-driven growth

📰 OpenAI News | 📅 July 21, 2025

OpenAI partners with the UK Government to boost AI adoption, drive economic growth, and enhance public services for a thriving AI ecosystem in the UK.

Thinking longer reduces AI models accuracy!

📰 Anthropic | 📅 July 19, 2025

Anthropic researchers: Longer reasoning in Large Reasoning Models (LRMs) can reduce accuracy, causing distraction, overfitting, spurious correlations, focus loss, and even problematic behaviors. Evaluations must test varying reasoning lengths to identify and mitigate these failures.