according to the financial times, meta is developing a highly personalized ai assistant for its more than 3 billion users, designed to help them accomplish everyday tasks. powered by the company’s new muse spark ai model, the assistant is currently undergoing internal testing by a team of employees. sources reveal that meta aims to create a product similar to openclaw, which allows users to build ai agents capable of autonomously performing tasks such as web browsing, email management, or calendar scheduling.
meta hopes users will voluntarily share highly sensitive information—such as health and financial data—with the assistant; however, insiders are skeptical, noting that “the trust gap is as wide as the grand canyon.” this move underscores ceo mark zuckerberg’s determination to deeply embed ai at the core of consumer products, even as investors grow increasingly concerned about soaring ai spending and the high costs of executing his ambitious vision. notably, meta plans to lay off 10% of its workforce later this month, yet zuckerberg continues to invest billions of dollars in ai infrastructure and talent acquisition.
the concept for these personal assistants was internally presented to employees at last week’s all-hands meeting. during the earnings call, zuckerberg stated that openclaw remains difficult for most users to adopt, and he aims to deliver a more refined, comprehensive, and immediately user-friendly experience. he acknowledged that while numerous agent-based tools are available on the market, “there are really only a handful that i would recommend my mother use.” however, openclaw has come under intense scrutiny due to security and privacy risks, particularly the potential for technological失控 if users grant access to their personal information. whether meta can strike a balance between powerful functionality and robust security will be pivotal to the success of its ai strategy.
