In this week’s AI, security researchers stole an artificial intelligence model with near-perfect accuracy by capturing electromagnetic signals.as Massachusetts Institute of Technology has unveiled a robotic system that can handle oddly shaped warehouse packages with an 80% success rate.
Meanwhile, three countries have begun regulating AI. google We’ve released Gemini 2.0, which is more autonomous. Recognized by technology leaders more severe Challenges lie ahead, as some AI models make dramatic leaps forward, while others hit roadblocks.
AI models are vulnerable to electromagnetic signal theft
north carolina state university Researchers demonstrated the following security vulnerabilities: AI systemachieves over 99% accuracy in extracting AI models by capturing electromagnetic signals from computer hardware.
The technology does not require direct access to systems and is raising alarms for major technology companies, including: OpenAI, human And Google is investing heavily in its own AI models. The findings highlight the growing cybersecurity challenges as companies increase their reliance on AI for competitive advantage.
MIT develops AI system for better warehouse robots
MIT researchers have created a new AI system, PRoC3S. warehouse robot Handle oddly shaped luggage and move through crowded areas more efficiently.
The system combines AI language models and computer vision to test actions in a virtual environment before taking them. In lab tests, it completed basic tasks like drawing shapes and sorting blocks with 80% accuracy.
This technology aims to enable robots to perform complex warehouse tasks. usually Requires human dexterity.
Three countries take different approaches to advance AI surveillance
Bipartisan U.S. House Select Committee Recommends Industry-Specific Policies AI monitoring It was Congress’ first comprehensive framework, rather than broad federal regulations.
Meanwhile, Malaysia has established a National AI Office to coordinate policy and development as it positions itself as a technology hub.
and, The UK has introduced a consultation on copyright reform to balance AI innovation with protection of creative industries.
The move reflects growing global efforts to establish an AI governance framework.
Google’s Gemini 2.0 promises more autonomous AI operations
Google releases Gemini 2.0an AI system designed to handle complex tasks across multiple platforms with greater autonomy.
The system powers projects like Astra for Android devices and Mariner for web navigation, and signals a shift away from command-based AI to more independent operations.
The main feature is a unified approach to processing different types of information, integrating the processing of text, images and audio, which previously required separate tools.
Signs of AI progress are mixed as big tech companies chart the future
A.I. We are delivering Contrasting signals are emerging, with some models making dramatic leaps forward and others hitting unexpected roadblocks.
Technology industry leaders acknowledge that while advances in AI are not stagnant, future advances face tougher challenges. As companies race to develop more practical AI tools, they are carefully weighing investment decisions against technology curves that ignore the simple characteristics of accelerating or slowing down.
For all of our PYMNTS AI coverage, subscribe to our daily AI newsletter.
