Meta Connect 2026: Muse Goes Wearable, What It Means for Enterprise
Meta Connect 2026: Muse AI Agent Goes Wearable, Mac Computer Use Arrives, and VR Glasses Debut
September 23, 2026. Less than three weeks after launching Muse, Meta’s personal AI agent, CEO Mark Zuckerberg took the stage at the company’s annual Connect event in Menlo Park to announce the next chapter: Muse integrated into wearable hardware, running autonomously on your Mac, and soon reachable from a pocket-sized device.
“The centerpiece of our vision for what we’re building is Muse,” Zuckerberg said. “In the coming years, I expect that Muse is going to grow into the personal superintelligence that billions of people around the world are going to use to accomplish their goals and improve their lives.”
Muse already reached the top spot on the US App Store within weeks of its launch, surpassing ChatGPT. Connect 2026 was Meta’s statement that this momentum was just the start.
For enterprise AI teams, the announcements carry several implications that go well beyond consumer wearables.
Muse Is Moving Off the Screen
The single most consequential announcement for how AI agents will operate in the future: Muse is coming to Meta’s AI glasses lineup.
Users will activate Muse by saying its name. From the glasses, Muse can guide personalized workouts, log meals, book appointments, get flight prices, help identify and buy products in view, and execute background tasks, all without the user touching a phone. Muse works in the background and checks back when complete.
This is not a voice assistant wrapper. Muse on AI glasses uses Muse Spark, Meta’s agentic model, with access to the same connector ecosystem available on desktop. The difference is the input modality: ambient, continuous, always-on. The agent can act on what the user is looking at, not just what they describe.
For enterprise use cases, the near-term signal is field workers: technicians, logistics staff, sales reps, and healthcare practitioners who need access to information and task execution without screen interaction. Hands-free agentic capability has been a goal in enterprise operations for years; wearable AI glasses with a capable backend model is the first credible commercial path to it.
Meta confirmed Muse will arrive on AI glasses “in the coming months” from the September 23 announcement.
Mac Computer Use: Walk Away and Let Muse Work
Meta also announced Muse computer use for Mac. Users can assign tasks, step away, and return to completed work.
“You can have it help you run your small business, or just get work done for you,” said Alexandr Wang, Meta’s chief AI officer. “You can walk away from your computer and it keeps working for you on all the jobs you lined up.”
Muse can drive any application on a user’s Mac with permission granted. This positions Meta directly in territory that Anthropic’s computer use tools and OpenAI’s GPT-6 Astra already occupy. The difference is that Muse’s computer use ships alongside a consumer agent with a free tier, making it more accessible to small teams and solo operators than enterprise-tier APIs.
For organizations already evaluating autonomous task execution, Muse for Mac is now a third option with a consumer distribution advantage: if employees adopt it personally, IT teams will face adoption pressure from the bottom up rather than top down.
Muse Gets a Digital Face, an Email Address, and More Connectors
Meta introduced three additional capability expansions at Connect that matter for enterprise planning:
Muse Realtime Avatar. A new AI model called Muse Realtime Avatar enables Muse to appear as a real-time, expressive, interactive avatar during video conversations. This is Meta’s first step toward giving agents a persistent visual identity, not just a voice.
Muse’s own email address. Muse will have an email address it can use to receive tasks, respond to correspondence, and participate in email threads. Users can forward emails to Muse or add it to a thread directly. This is a milestone: an agent that holds a communication identity within organizational systems can operate as a participant, not just a tool. For enterprise teams assigning delegated correspondence and asynchronous workflows to agents, this changes what “handing off to an agent” looks like in practice.
New connectors. Meta added Walmart, Best Buy, Sephora, Gap, Wayfair, Expedia (coming soon), and Instacart for commerce; GitHub, Notion, and Box for productivity. Payments now support Shop Pay and PayPal in addition to Stripe and Shopify. Wang noted that Meta opened connector development to third parties and received more than 1,500 applications in under one week. “I think this could be a generational opportunity to start building for a new platform as it takes off,” he said.
The GitHub and Box connectors, in particular, open Muse to enterprise use cases around code management and file operations, where Muse can take action within existing workflows rather than requiring a separate interface.
New Hardware: Audio Glasses, Gen 3 Ray-Bans, and VR Glasses
Connect 2026 also delivered three hardware announcements with enterprise relevance:
| Device | Price | Key Specs | Availability |
|---|---|---|---|
| Ray-Ban Meta Audio | $349 | 43 grams, 12hr battery, no camera, AI + audio only | Pre-order now, ships Oct 13 |
| Ray-Ban Meta Gen 3 | $449 | 12 MP camera, 3K Ultra HD, 9hr battery, 6-mic array | Available now |
| Meta VR Glasses | $1,299.99 | ~100 grams, 5K micro-OLED, belt-clip compute puck | Spring 2027 |
Ray-Ban Meta Audio addresses one of the core barriers to workplace adoption of AI glasses: the camera. By launching a camera-free, audio-only model at $349, Meta creates an option that sidesteps the privacy and policy concerns that have blocked camera-equipped glasses from many enterprise environments. The 12-hour battery and all-day comfort positioning (43 grams, adjustable nose pads and temple tips) make it viable for extended wear.
Ray-Ban Meta Gen 3 upgrades the camera to 12 MP and 3K Ultra HD video, improves battery to 9 hours, and adds a 6-mic array that cuts more than 90 percent of background noise. The new action button provides direct access to Muse. These improvements address the main friction points from Gen 2 in professional settings: audio clarity on calls and camera quality for documentation.
Meta VR Glasses are the most significant long-term signal for enterprise workspace design. At approximately 100 grams, they create virtual multi-panel workspaces on any surface. A puck clips to a belt or bag to offload processing and battery weight from the eyewear itself. Navigation happens through eye tracking and hand gestures. The Hologram feature enables photorealistic real-time video calls. Spring 2027 is the launch target at $1,299.99.
Private Processing is also coming to AI glasses, allowing local AI processing without data exposure to Meta’s servers. This feature addresses one of the principal barriers to enterprise adoption of cloud-connected wearables.
Enterprise Implications: Three Signals to Track
1. The agent-as-participant model is accelerating.
Muse’s email address is one signal. The Realtime Avatar is another. Agents acquiring communication identities within organizational systems, not just API endpoints, changes the security, governance, and workflow design requirements for enterprise AI programs. Governance frameworks built for tool-shaped AI agents will need to extend to cover persistent-identity agents.
2. The connector ecosystem is the real competitive moat.
1,500 developer applications in one week suggests the Muse connector platform is attracting serious developer attention. The addition of GitHub, Notion, and Box means Muse can participate in software development, knowledge management, and file operations workflows that enterprise teams already run. The platform dynamic, not the model, may be where Meta builds its deepest enterprise advantage.
3. Camera-free wearable AI is now commercially available.
Ray-Ban Meta Audio at $349 with 12-hour battery and no camera represents the first mass-market device that combines ambient AI access with a form factor acceptable in most enterprise environments. For organizations piloting wearable AI for field operations or knowledge work, this removes the policy blocker that has prevented broader trials.
What Happened and What It Means
Meta Connect 2026 was not primarily a hardware launch. It was a platform announcement. Muse moved from a mobile app to a platform that will run on wearables, desktops, glasses, pocket devices, and eventually headsets. The connector ecosystem grew. The agent acquired communication identity. The hardware expanded to fit more environments.
For enterprise teams building AI-native operations, the practical questions this week are:
- Does your AI governance framework account for agents that hold email addresses and participate in communication threads?
- Are your field operations pilots blocked by camera policy? If so, Ray-Ban Meta Audio arrives October 13.
- Is your team evaluating Mac computer use options? Muse is now a third path alongside Anthropic and OpenAI.
- Are you building on the Muse connector platform? 1,500 developers filed applications in the first week. That pipeline matters.
Personal superintelligence, as Zuckerberg framed it at Connect, is the long arc. The near arc, which arrived this week, is an AI agent that is ambient, participatory, and increasingly embedded in the hardware and communication channels where enterprise work already happens.
Primary sources: Meta Connect 2026 Official Announcement (meta.com, September 23, 2026); Muse on AI Glasses (meta.com, September 23, 2026); Ray-Ban Meta Audio Press Release (about.fb.com, September 23, 2026); Everything New Coming to Muse (TechCrunch, September 23, 2026); Meta Connect 2026 Coverage (The Verge, September 24, 2026).