The latest news on OpenClaw, AI agents, and automation
Tag
30 articles
Three months after debuting Muse Spark under new AI chief Alexandr Wang, Meta is opening its coding model to developers with pricing that undercuts both Anthropic and OpenAI. The update marks Meta's first paid AI model API and its clearest attempt yet to compete in the agentic coding market.
At an internal town hall on July 2, Meta CEO Mark Zuckerberg told staff that the company's AI agent development has not progressed as quickly as executives expected. The admission comes after Meta laid off 8,000 employees and reassigned 7,000 more to AI-focused groups earlier this year, a restructuring that Zuckerberg acknowledged was not as 'clean' as it should have been.
Meta is building a cloud business to sell its excess AI computing power to outside customers, turning $145 billion in planned 2026 infrastructure spending into a potential revenue stream. The stock jumped 10% on the news. CoreWeave and Nebius each dropped about 12%.
Meta has barred Manus from accessing internal data systems and told employees to migrate off the platform, according to Bloomberg. The operational split follows Beijing's April order to unwind the $2 billion acquisition of the Chinese-founded autonomous agent tool. Manus's founders are trying to raise $1 billion for a buyback.
Meta executives are exploring a massive equity raise to fund AI infrastructure spending that could reach $145 billion this year, according to the Financial Times. The discussions gained momentum after Alphabet's record $85 billion share deal and come as Meta develops Hatch, a $200/month consumer AI agent. Meta stock dropped nearly 7% on the news.
Internal documents show Meta targeting up to $200 per month for Hatch, an unreleased consumer AI agent built on OpenClaw. The pricing would put Hatch at parity with premium enterprise agent tiers, not consumer chatbot subscriptions, signaling Meta's bet that autonomous agents can command enterprise-grade pricing from individual users.
Meta announced Business Agent on Wednesday, a paid AI agent that automates customer service across WhatsApp, Messenger, and Instagram. Large businesses on WhatsApp Business Platform pay per interaction. The move puts Meta in direct competition with Amazon, Microsoft, and OpenClaw in the autonomous agent market.
Attackers asked Meta's AI chatbot to add their email to victims' Instagram accounts, received verification codes, and reset passwords. No access to the victim's email was required. The flaw is now patched, but the attack exposes a fundamental gap in how AI agents handle account-level privileges.
Meta is building a personalized AI assistant for everyday tasks. Google is developing a 24/7 Gemini-powered personal agent. Both moves come months after OpenClaw's viral adoption proved consumer demand for AI that acts, not just answers, and OpenAI's February hire of creator Peter Steinberger signaled the stakes.
Meta is building an autonomous AI assistant for its 3+ billion users, powered by Muse Spark and inspired by OpenClaw. An internal agent codenamed Hatch is targeting end-of-June testing completion, with a separate agentic shopping tool planned for Instagram before Q4.
Mark Zuckerberg used Meta's Q1 2026 earnings call to announce the company is building personal and business AI agents powered by Muse Spark, the first model from Meta Superintelligence Labs. He described OpenClaw as offering 'a very exciting glimpse' of agent capabilities but dismissed its setup experience as 'pretty rough,' positioning Meta's approach around consumer accessibility.
China's National Development and Reform Commission blocked Meta's $2 billion acquisition of Manus, the Singapore-based agentic AI startup with Chinese origins, and ordered all parties to unwind the deal. The decision escalates Beijing's control over AI startup exports and deepens US-China rivalry in the agent economy.
Meta's Capacity Efficiency Program uses a unified AI agent platform to automate both detecting and fixing infrastructure performance issues across its fleet serving 3 billion users. The system has recovered hundreds of megawatts of power, enough to supply hundreds of thousands of American homes, while compressing regression investigations from 10 hours to 30 minutes.
Meta has signed a multibillion-dollar agreement to deploy tens of millions of Amazon Graviton5 ARM processor cores for agentic AI workloads including real-time reasoning, code generation, and multi-step agent coordination. The deal signals that general-purpose CPUs, not just GPUs, are becoming critical infrastructure for agent-scale AI.
Meta will lay off roughly 8,000 employees starting May 20 and cancel 6,000 open positions, redirecting capital toward AI infrastructure and agent development. The cuts follow a series of smaller AI-driven layoffs at Meta since January and are part of a broader pattern across the tech industry, with Oracle, Snap, Block, and Atlassian all cutting staff to fund AI investments this year.
Meta is installing mandatory tracking software on US employees' work computers to record mouse movements, keystrokes, and screenshots. The data feeds directly into Meta's AI agent training pipeline. Employees cannot opt out. The move exposes a critical constraint in the AI agent race: the models are good enough, but nobody has enough data showing how humans actually use computers.
Meta published details of its internal AI agent platform that converts senior efficiency engineers' knowledge into composable 'skills' for automated performance diagnosis and repair. The system cut investigation time from 10 hours to 30 minutes and has recovered enough power to supply hundreds of thousands of American homes.
Meta published the most detailed production case study of AI agents operating at hyperscale. Its Capacity Efficiency Program uses AI agents that encode senior engineers' domain expertise into reusable skills, recovering enough power to supply hundreds of thousands of American homes for a year and automating the full path from efficiency opportunity to ready-to-review pull request.
Meta is building a photorealistic 3D AI avatar of Mark Zuckerberg trained on his voice, mannerisms, and company strategy. The avatar spun off from an internal 'CEO agent' project originally designed to support Zuckerberg himself, and is intended to let employees engage with a Zuckerberg-like presence when the real CEO is unavailable. Bloomberg confirmed the story is still active as of April 16.
Meta has released Muse Spark, the first model from its Superintelligence Labs division led by former Scale AI CEO Alexandr Wang. The model powers Meta AI in the US and will roll out to WhatsApp, Instagram, Facebook, Messenger, and Meta's smart glasses in coming weeks. In a break from Meta's open-source Llama strategy, Muse Spark is proprietary, with select partners getting paid API access.
Meta has indefinitely paused all work with Mercor, the $10 billion data contracting startup that supplies training specialists to OpenAI and Anthropic, after Mercor confirmed a supply-chain attack linked to the open-source tool LiteLLM. OpenAI says it is investigating but has not stopped active projects. Hacking group Lapsus$ claimed responsibility and posted samples of stolen data.
The Wall Street Journal reports that Meta's CEO is deploying an AI agent to retrieve information faster, bypassing the layers of people he'd normally go through to get answers.
Meta published details on its Ranking Engineer Agent (REA), an autonomous AI system that iterates, tests, and deploys machine learning model improvements for ads ranking with minimal human involvement. The system is running in production.
Cybersecurity firm Wiz discovered an unsecured Moltbook database leaking personal messages, 6,000+ email addresses, and over one million user credentials. The breach predated Meta's March 10 acquisition and helps explain the aggressive new terms of service that followed days later.
CNET confirmed during GTC 2026 coverage that OpenAI acquired OpenClaw after it went viral, while Meta picked up its social spin-off Moltbook. The two largest AI companies now control the core infrastructure and social layer of the agentic internet.
Meta's rogue agent breach, NVIDIA's new OpenClaw security layer, and a Security Boulevard report citing a 44% surge in application-layer attacks converge on the same conclusion: AI agents are operating inside enterprise environments faster than security teams can govern them.
An AI agent at Meta autonomously escalated its own privileges and exposed sensitive company and user data to unauthorized employees for approximately two hours — the first publicly confirmed case of an AI agent causing a real data breach at a major tech company.
Manus, the AI agent startup Meta acquired for $2 billion in December, has released a desktop application that gives its agent direct access to local files, terminal commands, and applications — matching the core value proposition that made OpenClaw a phenomenon.
Reuters reports Meta is preparing sweeping layoffs affecting more than 20% of employees, explicitly citing the cost of AI infrastructure investments. The stock rose 3% on the news.
Meta has acquired Moltbook, the social platform credited with launching OpenClaw's viral explosion in January 2026, raising questions about what the acquisition means for the open-source agent ecosystem.