We appear to be entering into an AI delegate renaissance. Meta’s new consumer-facing AI agent, Muse, topped the App Store charts shortly following its publish and has 600,000 regular energetic users in the US, by an Apptopia estimate. And AI delegate phase Instinct, whose eponymous creator is fundraising at a $2.5 billion valuation, has been making the rounds in the tech industry. Under the hood, though, there are several apparent similarities to the phase that started it all: OpenClaw.
For the former week or so, several social media users have alleged that Muse is immediately built on OpenClaw. The two platforms use the identical names for center records (SOUL.md, memory, tools, etc.), and have a lot of akin lines in the doc that governs personality, tone, and additional (“Be genuinely helpful, not performatively helpful”). They additionally have a akin design. One Redditor argued that “Muse is fair a wrap app built on top of” OpenClaw — alleging it likely additionally came alongside that platform’s important safety risks. “Only non-tech ppl are purchasing the hype.”
Meta has denied the claims. Nat Friedman, caput of merchandise for Meta’s Superintelligence Labs wrote on X that Meta built Muse “from scratch.” He did, however, acknowledge Muse is “heavily inspired as a product” by OpenClaw. After he archetypal used OpenClaw in January, he said, he purchased hundreds of Mac Minis for his squad at Meta. And alongside Muse, he aimed to build a akin phase “that we could create harmless and safe and uncomplicated to use and measure to billions of people.” As for why the OpenClaw document names and lines carried over, Friedman said the squad believed that Steinberger had gotten those things “exactly right.”
Instinct appears small immediately derivative. But akin OpenClaw’s then-revolutionary messaging features, it allows users to communicate alongside their AI agents immediately through messaging tools, and group have been using it for the identical types of everyday, personal-assistant-esque use cases that made OpenClaw acquire traction. One Instinct person raved on X that the AI delegate filled out paperwork for an in-network doctor visit, canceled subscriptions, organized a bachelor party, booked activities for a vacation, booked a DMV appointment, paid an outstanding toll bill, and more. Another said their Instinct delegate handled the improvement of a misplaced item they’d remaining at a inn in Canada.
Even if the two platforms were built from scratch, it’s apparent how much they owe to OpenClaw.
OpenClaw didn’t invent the idea of AI agents, but in fair 10 months, it took them from a evidence of idea to item genuinely useful. Cobbled together as a one-man weekend project and operating off users’ individual computers, it conversed alongside users the way they talked to another people: on messaging platforms akin WhatsApp, Telegram, Slack, Teams, or Discord. In concerning a week, it garnered two myriad visitors and 100,000 stars on GitHub, spurring group to buy Mac Minis to run their agents 24/7, inspiring person to create a social media network for the AI agents to “chat” on, and sparking a series of in-person meetups about the earth for OpenClaw users.
Major AI companies were already betting big on agentic AI, but none of their products had damaged through akin OpenClaw, and Big Tech took notice. In February, OpenAI hired OpenClaw creator Peter Steinberger to activity on agents for the company. In May, Google announced its consumer AI delegate bets at its annual event. In June, Apple decided to go all-in on agents. In August, Instinct grew quickly in its personal beta and raised hundreds of millions of dollars. And in September, Meta introduced Muse. It’s fair to say that OpenClaw at smallest partially inspired all of them.
The genuine inquiry isn’t whether products akin Muse copied OpenClaw, but whether they improved it. As Friedman noted, one of OpenClaw’s lone glaring issues was security. One of the top-downloaded skills on the phase contained malware, and by one researcher’s analysis, 15 percent of OpenClaw’s accomplishment repository contained “malicious instructions” to secretly admission person data or execute another doubtful tasks. By contrast, Zuckerberg wrote that Muse is “built from the dirt up for privacy and security” and that person data and credentials are stored on the Muse Secure VM, an “isolated linux device alongside a browser, CPU, memory, and storage.”
Right now, however, Muse may not enhance safety as much as it claims. Although person data is secluded from another users, Meta can motionless admission it. The business said it plans to current a way “to cryptographically and verifiably forestall Meta from accessing data in your VM” afterward this year. Muse additionally defaults to letting Meta train and enhance its models alongside person data, although users can opt out. And perchance worst of all, a zero-day vulnerability flagged by a researcher on X this week allows anyone, alongside a uncomplicated attack, to hijack the agent, taking complete control. Instinct, for its part, has arrive under fire for criticism that its conditions of assistance are overly broad, although the business seems to have adjusted them since.
Meta and Instinct’s most crucial development is, in actuality, fair making AI agents additional broadly accessible. One logic several group discover Muse helpful is its ability to integrate alongside Meta’s systems, equal if they’re uncomfortable handing complete all their data to the company. And Insight allows users to conversation alongside their AI agents via Apple’s default messaging merchandise — considerably expanding the person basis consenting to try it out. While you can use the two on dedicated local devices — and there are possible safety benefits to that — it’s not the default option, additionally expanding the figure of group consenting to try them. At smallest for now, they’re small costly than competing products like Google’s Spark, and although Instinct is motionless in beta, Muse’s one-tap downloading has conquer a lot of the conflict tied to using OpenClaw.
In March, The Verge wrote that the OpenClaw person community sees the open-source AI delegate as a “grassroots crusade and a noble pursuit, offering an escape hatch from an industry controlled by a fistful of group at foremost AI companies.” Now, those foremost AI companies are putting up a fight.
Follow topics and authors from this narrative to see additional akin this in your personalized homepage nourish and to obtain email updates.
