Newsfeed

Tech News Digest: Uber AI Caps, Amazon Proteus & AI Security Models

Uber AI Costs: Rideshare Giant Caps Employee Usage of Claude Code

To manage rising Uber AI costs, the rideshare giant has implemented strict spending caps on artificial intelligence coding tools for its workforce. Uber is now limiting individual employee expenditures to $1,500 per month for AI assistants such as Claude Code.

This usage cap highlights the growing financial challenges of enterprise artificial intelligence adoption. Companies are increasingly forced to balance the productivity benefits of generative AI with the substantial subscription and compute expenses required to maintain these advanced developer tools.

Robotics & Autonomous Vehicles: Amazon AI Proteus Robot Launch and US GUARD Act

Amazon has introduced an upgraded, fully autonomous Proteus warehouse robot equipped with advanced AI that allows workers to issue plain-language commands. The next-generation system can navigate entire fulfillment centers to move heavy carts and will deploy across Europe in the first half of 2027 as part of a €10 billion investment.

Concurrently, US lawmakers have proposed the bipartisan GUARD Act of 2026 to evaluate the national security risks of foreign-made humanoid and quadruped robots. Under this legislation, autonomous systems from countries of concern that pose unacceptable threats will be restricted via the FCC’s Covered List. These developments highlight both the rapid commercial advancement of autonomous vehicles and the increasing regulatory scrutiny surrounding connected robotics.

Sources:

Generative AI Performance Reports Launched in Google Search Console

Generative AI performance reports have officially launched in Google Search Console. Google designed this new SEO tool to provide site owners with dedicated views of their visibility within generative AI search features.

The AI reports specifically track impressions across:

  • AI Overviews
  • AI Mode
  • Generative AI elements in Google Discover

Previously bundled into overall metrics, this standalone data empowers webmasters to analyze specific URLs, geographic breakdowns, and device usage for AI-driven traffic. This rollout establishes a critical measurement framework for organic search performance as generative AI platforms reshape information retrieval.

Sources:

AI-Driven Vulnerability Discovery: Anthropic and Microsoft Unveil New Security Models

In spring 2026, the software engineering and open-source communities are experiencing a surge in AI-driven vulnerability discovery and security reporting.

Anthropic recently expanded Project Glasswing, an initiative utilizing its Claude Mythos AI model, to approximately 150 new organizations after discovering thousands of high and critical severity vulnerabilities across the open-source ecosystem.

Concurrently, Microsoft introduced Codename MDASH, an agentic vulnerability discovery system that orchestrates over 100 specialized AI agents to find and validate exploitable flaws. These new security models are already proving effective, as MDASH has discovered and patched 16 new vulnerabilities across the Windows networking and authentication stack.

Industry experts at Red Hat note that this rapid influx of AI-discovered vulnerabilities—including recent Linux kernel flaws like Copy Fail and Dirty Frag—underscores the growing need for human technical advocacy and robust security verification.

Sources: