All the doom and gloom scenarios surrounding AI becoming ever more powerful seems to have taken a back seat, with the launch of Meta AI’s Muse and now OpenAI’s Dots, in case you didn’t notice. It’s time for AI to become less threatening and full of optimism. That’s at least the key takeaway from recent announcements from Meta and OpenAI.
Also read: This Indian AI startup wants to disrupt video production workflows: Here’s how
Muse was revealed at Meta Connect 2026, after being announced earlier in September, with a very cute agent mascot called Jolly. Where the AI overload truly became clear was when Meta showed off a Realtime Avatar model, which allows video calls with an animated version of the agent in action.
Beaten to the opening punch, but not waiting too long for its first salvo, OpenAI decided to go further in the friendliness spin to AI with Dots. It’s no terminator, but a primary dot, where users get to do everything from naming, personalizing the journey and making it their own. OpenAI’s launch material even showed off a blue dot named Felipe, which looked like a cartoon doodled, nothing too scary.
And early indications show that people seem to like it, at least according to Sensor Tower. Muse shot to the top of the free app chart on Apple’s App Store 10 days after its September 8 debut, passing ChatGPT, and was downloaded more than 2.5 million times since launch. And Meta stock climbed as much as 25% in September, much of that since the Muse launch.
Also read: GPT-6.1 Sol: Everything that’s better than the week old GPT-6 Sol
The early narrative around these two AI assistants with agentic use cases is far away from all the fear and anxiety that’s been surrounding the industry all of 2026. Not panic, no scary business, Muse and Dots are offering something refreshingly mundane. Muse is capable of doing everything from booking a date night to emailing a friend about an upcoming road trip that it planned on your behalf.
Built on GPT-6 Astra, Dots has a cloud computer squeezed inside an ordinary browser tab, can work on delegated tasks 24/7 and connects to over 4,000 apps. Suddenly, the pitch isn’t superintelligence. It’s getting your errands done with an overdose of cuteness, might I add.
Both Meta and OpenAI are also leaning hard on the language of controls available to users of Muse and Dots all the time. Dots follow ChatGPT’s permissions and support custom rules for which tasks require approval. According to Meta, Muse doesn’t share conversations or virtual machine session data with its ad systems, as of now.
You’d do well to know that beneath the surface-level cuddly avatars sits technology that’s arguably more autonomous than anything Meta and OpenAI have made available to consumers before. How safe are their respective AI agent guardrails enforced remains to be seen, to be honest.
There’s a commercial undercurrent here, don’t forget. Meta’s 2026 capital expenditures are in the range of hundreds of billions of dollars, while OpenAI has unveiled a new $500-a-month ChatGPT subscription plan. In the race to monetization and chasing IPO, Muse and Dots seem like a business necessity. Not doom and gloom, but friendly faces of Muse and Dots is what will move the needle favourably in the acceptance of always-on agents in web browsers, apps, and social media networks.
So has AI actually become less dangerous? Not quite. The debate on how to maximise AI benefits without endangering humanity rages on. For now, it’s been given a softer, friendlier face.
Also read: OpenAI Dots explained: An AI coworker that never logs off