發生了什麼事
The Verge reports that social media users have alleged Meta's new consumer , Muse, is directly built on the open-source platform OpenClaw due to identical file names and personality documentation lines. Meta's head of product for Superintelligence Labs, Nat Friedman, denied that Muse was built on OpenClaw, stating it was built from scratch but heavily inspired by it. The report also highlights that Instinct, another AI agent platform, shares similar messaging-based features with OpenClaw. While Meta claims Muse is more secure, the article notes that a zero-day vulnerability was recently flagged in Muse, and Meta retains access to user data in its isolated VMs.
The Verge reports that allegations have circulated on social media suggesting Meta's Muse is a wrapper around the open-source platform OpenClaw. Evidence cited includes identical core file names such as SOUL.md and memory, as well as matching lines in personality-governing documents, such as the instruction to be 'genuinely helpful, not performatively helpful.'
Nat Friedman, head of product for Meta’s Superintelligence Labs, responded to these claims on X, stating that Meta built Muse 'from scratch.' However, he acknowledged that Muse is 'heavily inspired as a product' by OpenClaw, which he first used in January. Friedman explained that the team purchased hundreds of Mac Minis to test the platform and aimed to create a version that could be made safe, secure, and scalable to billions of users.
The report also examines Instinct, an platform fundraising at a $2.5 billion valuation. Like OpenClaw, Instinct allows users to communicate with agents via messaging tools. Users have reported using Instinct for tasks such as filling out medical paperwork, canceling subscriptions, and booking appointments, mirroring the use cases that drove OpenClaw's initial traction.
The Verge notes that while OpenClaw was a one-man weekend project that gained massive popularity, it had significant security issues, including malware in its skill repository. Meta claims Muse addresses these issues by storing user data in an isolated Linux VM called the Muse Secure VM. However, the article points out that Meta can still access this data, and a zero-day vulnerability was recently identified that allows attackers to hijack the agent.
為什麼這很重要
This development clarifies the lineage of major commercial AI agents, confirming that OpenClaw's significantly influenced industry leaders like Meta. It highlights the tension between rapid product adoption and security integrity, as both Muse and OpenClaw face criticism for security flaws. The report underscores that while commercial agents offer better accessibility and integration, they may not yet resolve the fundamental security risks present in the underlying open-source architecture that inspired them.
The confirmation that major AI products are derived from or heavily inspired by OpenClaw validates the impact of open-source development on the broader industry. It suggests that the 'grassroots' approach of OpenClaw has become a blueprint for commercial offerings from major tech companies.
The security implications are significant. While Meta positions Muse as a secure alternative to OpenClaw, the persistence of a zero-day vulnerability and the company's retained access to user data raise questions about the actual security improvements. This is critical for users who may be adopting these agents for sensitive personal tasks.
The report highlights a shift in the market from experimental, local-first tools to scalable, cloud-integrated commercial products. This shift brings broader accessibility but also concentrates control and data handling in the hands of large corporations, potentially altering the privacy landscape for AI users.
互動機制：它實際上是如何運作的
以互動方式探索這項發展背後的基礎技術。
crm_get_transaction(id='4092').What most distinguishes an AI agent from a basic chatbot?
接下來看什麼
Monitor for independent security audits of Muse's isolated VM architecture to verify Meta's claims about data protection. Watch for further regulatory or public scrutiny regarding the similarity between Muse and OpenClaw, as well as the implementation of Meta's promised cryptographic measures to prevent company access to user data later this year.
Independent verification of the security claims surrounding Muse's isolated VM environment is needed, particularly regarding the zero-day vulnerability mentioned in the report.
The implementation of Meta's promised cryptographic measures to prevent company access to user data later this year will be a key indicator of whether Muse truly improves on OpenClaw's security model.
Further developments in the legal or public discourse regarding the similarity between Muse and OpenClaw, as well as the broader implications for open-source adoption by major tech firms.