OpenAI, Meta, and Anthropic — three of the biggest names in AI — are giving their AI agents cartoon mascots. According to U.S. media reports this week, the companies hope friendlier characters can make software that sends emails, books appointments, and reads files on a user's behalf feel more like a companion, winning over ordinary users who are still wary of AI.
Each company has its own
OpenAI has gone furthest. Dots, its personal agent unveiled at the September 29 developer conference, runs on GPT-6 Astra and first rolled out to ChatGPT Pro and Business Premium users in eligible markets, with Europe and the UK left out for now. In the demo, every Dot has its own character: Alfred is a yellow pear with glasses and a bow tie, Felipe is a blue cloud in a beret, and the default character, Dottie, is a green bean with two diamond-shaped eyes. OpenAI says the characters were designed with an outside partner and were partly inspired by a pet feature previously built into Codex, with the goal of feeling "friendly, not intrusive."
Meta's Muse launched on September 8 and later added a character named Jolly — plush-toy in texture, with rosy cheeks — whose appearance users can customize themselves.
Anthropic's mascot is Clawd, a pixel-art orange crab that has mostly appeared in the Claude Code terminal interface.
Clippy's cautionary tale
Using a cartoon mascot to soften a tech product isn't new. Microsoft put the animated paperclip Clippy into Office, and it went on to become shorthand for "annoying assistant"; today Microsoft pairs Copilot with Mico, a smiling blob-shaped character. Apple, Mozilla, and Reddit have their own mascots too, and Duolingo's green owl has more than 20 million followers on social media.
The difference is what these agents can actually do. Clippy only popped up with suggestions while you were writing a letter; Dots and Muse need access to email, calendars, files, and even payment permissions to act on a user's behalf. Giving the latter a round, friendly face lowers the psychological barrier before someone clicks "allow" — it doesn't shrink the scope of what's being authorized. This week alone, controversy over Muse reading private information on Mac computers prompted Apple to announce it would tighten macOS's full disk access permissions.
Ed Zitron, founder of the PR firm EZPR, put it bluntly:
"It's entirely marketing. It exists to get quoted in articles that say that 'Meta has given AI a cute face.'"
Justin Joseph, an associate professor of public relations at Boston University, argues anthropomorphizing has a real function — people are more comfortable talking about "my Roomba" than "my autonomous robotic vacuum." Anthony Patterson of Lancaster University in the UK offered a counterpoint in an interview this past May: these mascots will eventually talk to users one-on-one and try to persuade them to do things, which he called "kind of creepy."
Chinese tech firms did this years ago
For Chinese users, this playbook is nothing new. Voice assistants from domestic phone makers, customer-service bots on e-commerce platforms, and many large-model apps have long paired AI with a fixed cartoon mascot and name. The difference is that those earlier characters mostly sat behind questions and recommendations, while this latest wave of mascots abroad belongs to agents that can directly take action.
Dots has not entered the Chinese mainland market, and Muse's initial launch was U.S.-only. Clawd, however, shows up in the terminals of plenty of domestic developers through Claude Code. Beyond the characters themselves, none of the three companies has published data on whether users make different permission decisions when facing a cartoon mascot versus a plain chat window.
Sources: The Wall Street Journal, TechCrunch, CocoLoop, Fortune, and AOL; Dots' rollout scope and character design follow OpenAI's developer conference demo, and expert views are drawn from public interviews.