Anthropic previews Claude Duet, a proactive agent that works in the background.
Anthropic has introduced Claude Duet, a separate app in which Claude works as a proactive agent in the background. Launched on 21 October as a research preview, it is open by waitlist, starting with Pro and Max subscribers in the US.
Built on Claude Opus 5.5, Duet sends a short morning brief on what matters across a user's connected apps, checks those accounts every few hours and gets in touch only when something needs attention. It takes on tasks from scheduling meetings to building dashboards, forecasts and prototypes, running on its own virtual machine with its own browser, so the user's laptop or phone does not need to stay on. It carries over memory, skills and connected apps from the user's Claude account and can draw on more than 6,000 connectors and plugins, including Gmail, Slack, HubSpot, Canva, LinkedIn and Shopify.
Anthropic says Duet shows the user exactly what it plans to do before it sends messages, makes payments, signs documents or deletes anything, and it cannot trade securities or move money between accounts. It can buy through Link, Stripe's digital wallet, using a virtual card, and check out with Shop Pay. The company is plain about its limits: it may misread a thread or misjudge urgency, it is less reliable on personal tasks, and background checks and large jobs count against plan usage limits. Duet is included in the Pro and Max plans and comes as standalone iOS and Android apps and on claude.ai.
Anthropic has not said when Duet will reach users outside the US, and it says more detail on the safety design will follow in the coming days. It is not yet open to agencies and founders in the Gulf, but it shows where AI assistants are heading: from answering prompts to watching inboxes and calendars and acting with approval. Teams planning to use such agents will need clear rules on what an assistant may send, buy or sign on their behalf.
Why it matters
AI assistants are moving from answering prompts to acting on their own, with approval. Gulf teams should set rules now on what agents may send, buy or sign for them.
Related trends.
Zuckerberg says shopping is the top use for Meta's Muse agent
Meta's Muse AI agent has been downloaded more than 10 million times since its US and Canada launch last month, and shopping is its most common use. Mark Zuckerberg said average purchases through Muse are above $200.
Why it mattersMeta wants its AI agent to sit between shoppers and brands, which would shift commerce from websites to agent-to-agent deals.
ChatGPT Ads adds 26 countries, but names no Middle East market
OpenAI has opened ChatGPT advertising in 26 more countries, bringing the total to 89. It says more than 100,000 advertisers have run campaigns since ads launched just over eight months ago.
Why it mattersChatGPT is becoming a buyable channel in most major markets, but OpenAI has not said which Middle East countries are now live.
One in three Saudi and UAE online adults research purchases with AI
More than one in three online adults in Saudi Arabia (33.8%) and the UAE (33.4%) say they use AI tools to research potential purchases, DataReportal's Digital 2027 data shows. Worldwide the figure is 26%, which puts both Gulf markets well above the global level.
Why it mattersA larger share of online adults in Saudi Arabia and the UAE research purchases with AI than the global average. Brands should check what assistants say about their products, in Arabic and English.
