Meta Launches Muse Charm AI Device, Beating OpenAI and Google

Key Takeaways

• Meta launched the Muse Charm on September 23, 2026, about a week before OpenAI's expected entry into consumer AI hardware.
• Meta stock rose 4.5% on announcement day and over 25% since the Muse Charm reveal.
• The Charm is a keychain-sized device that ships in December and works with Instagram and WhatsApp for AI agent tasks.
• Muse will also come to Meta's AI glasses in the coming months, with Realtime Avatar previewed for video and speech sync.
• Trump renamed artificial intelligence as Super Intelligence after meeting with Chinese President Xi Jinping.
• Trump and Speaker Johnson will meet AI executives at the White House on September 29 amid pressure to regulate AI before midterms.
• Policy-as-Skill framework achieved 53.8% accuracy on 600 tasks, improving to 61.2% with deterministic control.
• IndicBankBench found AI banking reliability between 43.7% and 58.2% across 11 models, questioning one-time success metrics.
• Ambassador Dorothy Shea says innovation and regulation are not mutually exclusive in AI policy.
• FutureX Capital founder Cynthia Zhang sees strong potential in China's AI ecosystem but warns of bubble-like valuations in robotics.

Meta launches Muse Charm AI device ahead of rivals

Meta CEO Mark Zuckerberg introduced the Muse Charm, a new consumer AI device, to beat competitors like OpenAI and Google. Investors reacted positively, pushing Meta stock up 4.5% on Thursday and over 25% since the announcement. This move follows years of struggle for Meta in the virtual reality market after changing its name five years ago. The company hopes this new hardware strategy will finally succeed where previous attempts failed.

Meta releases Muse Charm to beat OpenAI in AI race

Meta launched the Muse Charm, a standalone AI companion device, to enter the consumer AI hardware market before OpenAI and Google. This launch marks Meta's biggest hardware bet yet as it tries to own the next computing platform. The device focuses on AI agents rather than simple features, aiming to compete with struggling startups like Humane and Rabbit. While previous hardware like Portal and Ray-Ban glasses had limited success, growing familiarity with AI assistants may help this new product succeed.

Meta Muse Charm aims to dominate AI agents market

Meta released the Muse Charm to enter the consumer AI hardware market before OpenAI launches its own devices. The device works with Instagram and WhatsApp to let users schedule posts or manage messages through an AI agent. This approach differs from competitors like Humane and Rabbit by focusing on AI agents instead of just AI features. Success depends on convincing consumers that dedicated AI hardware offers better benefits than using AI on smartphones.

Meta unveils Muse Charm and AI glasses plans at Connect

Meta showcased its Muse AI agent at the Connect 2026 conference on September 23, revealing the new Charm gadget and plans for AI glasses. The Charm is a small keychain device scheduled to ship in December, designed to let users talk to Muse when they are not wearing glasses. Muse will also appear on Meta's AI glasses in the coming months, allowing the agent to act on what users see. The company also previewed Muse Realtime Avatar, which pairs generated video with synchronized speech for a better user experience.

Meta Charm device arrives one week before OpenAI

Meta launched the Charm device on September 23, 2026, entering the consumer AI hardware market one week before OpenAI. This pocket-sized gadget runs the Muse AI agent and is separate from Meta's Ray-Ban and Oakley smart glasses lines. No pricing was announced at the Connect conference, but the company plans to ship units during the 2026 holiday shopping season. Analysts view this as an unproven strategy because Meta reportedly built only a small number of working units before the announcement.

Trump and Johnson to meet AI executives at White House on September 29

President Donald Trump and Speaker Mike Johnson will meet with top AI company executives at the White House on September 29. The meeting comes as Republicans face pressure to regulate AI ahead of the November midterm elections. Johnson has resisted quick regulation, saying the House will not rush with knee-jerk rules. Leaders argue that heavy regulation could hurt U.S. national security and give China an edge in AI development.

Trump rebrands artificial intelligence as Super Intelligence

President Donald Trump has renamed artificial intelligence as Super Intelligence. He announced the rebranding on social media Friday after meeting with Chinese President Xi Jinping. Trump said the new name is more accurate and important. Several tech CEOs including Sam Altman, Mark Zuckerberg, Jensen Huang and Sundar Pichai attended a state dinner with Xi on Thursday night.

Policy-as-Skill framework improves governed AI decision making

Researchers introduced Policy-as-Skill, a modular system that helps organizations use AI for policy and compliance decisions. It packages evidence validation, review routing and audit functions into versioned tools. Tested with a Gemma4 backend on 600 tasks, PaS+Audit achieved 53.8 percent exact accuracy and perfect audit completeness. Adding deterministic control raised accuracy to 61.2 percent but results varied by task.

IndicBankBench tests AI assistants in Indian retail banking

Researchers created IndicBankBench, a 799-case benchmark for testing AI assistants in Indian retail banking. It covers five operational domains and evaluates safety, tool use, response quality and advisory accuracy. Across eleven models, strict reliability ranged from 43.7 percent to 58.2 percent, while at-least-once success ranged from 60 percent to 74 percent. The findings show that one-time success can overstate how dependable AI banking behavior really is.

FutureX Capital founder highlights China AI growth and robotics risks

Cynthia Zhang, founder of FutureX Capital, discussed opportunities in China's AI ecosystem including vertical applications and open-source models. She also warned that parts of the physical AI and robotics market already show bubble-like valuations. Her comments came in a CNBC interview on September 25. Zhang sees strong potential in AI but urges caution in robotics investments.

Citrix launches AI browser session insights for agents

Citrix launched Session Insights for Citrix SecurAccess with Chrome Enterprise. The tool creates visual records of browser sessions for both employees and AI agents. It uses AI analysis to flag risky behavior and recommend security policy changes. Administrators can review how authority was used during workflows. The feature helps determine whether an action was taken by a worker or by software acting on their behalf.

SpaceX to launch Google AI satellite prototype next week

Google announced that a prototype satellite for Project Suncatcher will launch next week on Transporter-18. The satellite was built by Planet and carries four Google tensor processing units. It will test how these AI chips perform in Earth orbit for a year. Cooling remains a major challenge for orbital data centers because space lacks airflow to remove heat.

Maryland lawyer scolded for fake AI-generated citations

A Maryland judge reprimanded Donald Quinn of Quinn Patton for using fabricated citations in a court filing. Judge Charles C. Peters called the citations clearly fabricated and said Quinn abused the judicial process. The firm previously faced a lawsuit over AI-related errors in a case. The incident raises concerns about AI use in the legal profession and the risks of fake court citations.

3 AI-led stocks with strong estimate revisions and upside

Three AI-focused stocks are highlighted as investment opportunities: PWR, KEYS, and APH. PWR provides electric power solutions that help utilities optimize operations. KEYS offers electronic design and test automation tools for engineers. APH supplies interconnect solutions that help manufacturers cut costs. All three companies are benefiting from rising demand for AI and strong earnings estimate revisions.

Ambassador says AI innovation and regulation need balance

Ambassador Dorothy Shea says the debate over AI does not have to choose between innovation and regulation. She says the focus should be on weighing America's innovation advantage against concerns about AI agents acting independently. The comments highlight growing global attention to how autonomous AI tools should be managed.

Sources

• Meta gets to consumer AI device market before OpenAI, but Zuckerberg's strategy remains unproven
• Meta Beats OpenAI to Market with Muse Charm AI Device
• Meta Beats OpenAI to Consumer AI Hardware with Muse Charm
• Meta puts Muse at the center of Connect with Charm and AI glasses
• Meta Charm Beats OpenAI to AI Devices by 1 Week
• Trump and Johnson meeting with AI executives set for Tuesday
• Trump’s newest rebranding project: The technology formerly known as AI
• Policy-as-Skill: Governed LLM Decision Support with Evidence, Deterministic Control, and Audit
• IndicBankBench: Evaluating Safety and Reliability of Language Model Assistants in Indian Retail Banking
• FutureX Capital founder sees China AI opportunities, warns of robotics bubble
• Citrix adds AI browser session insights for agents
• SpaceX launching prototype Google AI satellite next week
• MD lawyer scolded for fake citations after past AI issues
• Buy 3 AI-Led Stocks Amid Solid Estimate Revisions and Upside Potential
• Innovation or regulation on AI?