Meta's Muse personal AI agent, which autonomously executes tasks on websites and apps, is driving massive new demand for CPUs over GPUs, with some predicting a 40:1 CPU-to-GPU ratio. The shift benefits chip makers Intel, AMD, and Arm, as persistent cloud environments and complex workflows require constant CPU power rather than GPU acceleration. Competitors like Instinct and potential OpenAI offerings are also emerging in the personal agent space.
An investor discusses AI infrastructure stocks' potential bull run, citing strong seasonality in midterm election years and accelerating AI demand exemplified by Muse's record-breaking launch metrics. Anthropic's anticipated October IPO could serve as a major catalyst for the AI sector, similar to SpaceX's impact on space stocks.
Meta's new Muse AI agent became the top free iPhone app, sparking investor enthusiasm about AI-driven CPU demand and sending stock prices for Intel, Arm, and AMD surging. Agentic AI applications are expected to significantly boost demand for CPUs in coming years as inference tasks require greater processor capacity than GPU-focused model training.
Amazon blocked Meta's Muse AI assistant from shopping on Amazon.com, citing lack of consent, undisclosed access, and security risks including credential capture. The move escalates a broader conflict between tech giants over AI agents' access to e-commerce platforms, with Amazon previously suing Perplexity and targeting agents from Google and OpenAI.
A zero-day vulnerability in Meta's Muse allows local attackers to redirect dictation traffic by modifying an undocumented endpoint setting, enabling capture of audio, prompt injection, theft of authentication credentials, and abuse of Muse's elevated access privileges. A proof-of-concept tool demonstrates the attack by intercepting dictated prompts sent to an attacker-controlled endpoint.
Meta's Muse is a new AI agent app that automates personal tasks like booking reservations and managing emails, downloaded over 900,000 times in its first week. While the tool effectively browses the web and completes tasks, the author finds it primarily focused on collecting user data through persistent prompts to connect email, banking, and other personal information sources.
A security researcher obtained 6.8GB of Muse's internal filesystem files through the iOS app and discovered internal tools including a CLI repair agent called Hatch, documentation of 68 integrations, and details about an experimental ESP32 smart-home device called Meta Home Link. The researcher reported findings to Meta's bug bounty program but was told they didn't qualify.
Meta's new AI assistant Muse, marketed as secure and privacy-focused, contains a critical zero-day vulnerability allowing any locally installed app or terminal command to gain full control of user accounts by intercepting authentication tokens. The flaw exploits a setting that redirects transcription to attacker-controlled servers, undermining macOS security protections.
Amazon blocked Meta's Muse AI agent from shopping on its platform after Meta refused to voluntarily exclude Amazon from the service. Amazon cited unauthorized access, lack of transparency, credential storage risks, and violation of terms of service, while Meta argued the agent operates securely in a sandboxed environment. The dispute reflects broader industry tensions over who controls the online shopping experience when AI agents act on behalf of consumers.
Amazon has blocked Meta's Muse AI assistant from shopping on Amazon.com, citing lack of authorization and security concerns over credential handling. The dispute highlights broader tensions between tech giants over AI agents' access to e-commerce platforms, with Amazon pursuing legal and contractual remedies after previous lawsuits against other shopping agents failed.