Skip to content
AI News

Tencent Releases Hy4 Model as Anthropic Builds FlashProve Identity System

Tencent Hy4 Anthropic FlashProve Claude Google Workspace for Education data centers Russia Ukraine

Tencent Holdings released its Hy4 open-weight foundation model on Wednesday, claiming it joins the top tier of Chinese open-weight AI systems for the first time, according to Bloomberg. The company also plans a Flash version of the model, marking a significant competitive push into the Chinese AI landscape dominated by domestic rivals.

Anthropic is developing a system called FlashProve that would let anyone verify an AI response genuinely came from the model it claims to be from, reports The Verge. The identity verification system addresses growing concerns about AI model impersonation and could become a critical trust layer for enterprises deploying Claude and other models in production environments.

Data center builders are turning to once-discarded jet engines as backup power sources amid years-long waits for grid access and new turbines, according to Bloomberg. The unconventional solution is less efficient and less environmentally friendly but highlights the desperate infrastructure crunch facing the AI industry as computing demands surge.

Russia is rushing to build underground data centers after Ukrainian drone strikes exposed vulnerabilities in its digital infrastructure, per Reuters. The hasty construction plans reveal how physical security has become a pressing concern for data center operators in conflict zones, with the digital infrastructure war backfiring on Russian forces.

Google will offer schools AI tools designed to not give away answers, but schools that opt out of AI features must also give up Google search, according to The Verge. The bundled approach affects millions of students using Google Workspace for Education and raises questions about competitive bundling in the edtech market.

The 2026 midterm elections will shape tech industry regulation over President Donald Trump's final 2 years in office, with affordability issues dominating races but several policy topics directly impacting the technology sector, reports Politico. The outcomes could influence AI governance, antitrust enforcement, and infrastructure policy through 2028.

Sources:

Enjoying this? Explore the podcast.

The AI Decision Course: 20 courses and 271 short lessons for directors, VPs and executives. One lesson from each course is free to read.

Share this post: