TREE NEWS update: Meta’s chief financial officer said the company is confident that artificial intelligence will become an important source of revenue diversification over time, with the new AI business expected to begin contributing incremental revenue starting in 2027. The CFO did not provide further financial details or specify the scale of the expected contribution.
TREE NEWS update: Meta’s chief financial officer said advertiser adoption of the company’s video generation feature rose 25% quarter-over-quarter. Meta One has surpassed 15 million trials and subscriptions to date, and Meta sees a path to monetizing Muse through subscriptions. The CFO also flagged a “major opportunity” in small businesses, adding that Muse’s compute intensity means expanding its user base will require significant investment and is expected to raise total cost of service, while compute efficiency is projected to improve 10-fold over the next year.
TREE NEWS reports: Google has stopped using nondisclosure agreements for its data centers. The change means the company will no longer require confidentiality pacts tied to those facilities. No further details on the scope or timing of the policy shift were provided.
TREE NEWS update: Meta Platforms CEO Mark Zuckerberg said glasses sales are continuing to grow rapidly and that with Muse the company expects paid subscriptions to increase. He also said Meta will adjust the pace of its virtual reality investments to focus on superintelligence.
TREE NEWS update: Mastercard has opened its Agent Pay for Machines pilot to developers, adding buyer- and seller-side APIs, a toolkit and a sandbox for testing. The program, which lets autonomous AI agents transact on behalf of users, remains a noncommercial pilot. The developer tools are meant to extend testing of machine-to-machine payments ahead of any commercial rollout.
TREE NEWS update: Meta’s board approved the creation of a publicly listed non-voting Class C common stock, with charter amendments effective January 19, 2027, the company said. Meta will issue one Class C share for each Class A and B share held as a one-time stock dividend, with record and distribution dates still to be set by the board. After distribution, the Class A share price is expected to reflect a two-for-one split effect. Meta also signed a founder alignment agreement with CEO Mark Zuckerberg covering his sale of the Class C shares, his service and future Class C dividends.
TREE NEWS update: Meta reported third-quarter revenue of $64.6 billion, up 26% year on year, while operating income fell 37% to $12.99 billion, dragged down by $9.9 billion in legal costs. Net income rose 264% to $9.86 billion, or $3.83 per diluted share, and daily active people averaged 3.64 billion in September, up 3%. Meta guided fourth-quarter revenue to $73–76 billion and narrowed its 2026 capital expenditure outlook to $135–145 billion.
TREE NEWS update: Nvidia introduced its OpenShell Reference Architecture, pairing a software sandbox on Vera CPUs with Sentry monitoring on BlueField-4 chips to let companies run autonomous AI agents while enforcing limits agents cannot reach. The blueprint is designed to contain and monitor AI agents as they operate.
TREE NEWS update: Meta Platforms now expects full-year capital expenditure of $135 billion to $145 billion, narrowing the top end from its prior $130 billion-$145 billion range. The company guided fourth-quarter revenue at $73 billion-$76 billion, against analyst expectations of $73.57 billion. It also disclosed a plan to issue a new class of non-voting Class C common stock.
TREE NEWS update: US Treasury Secretary Scott Bessent told investors that there is no systemic risk from artificial intelligence, pushing back against concerns that AI could destabilize financial markets. The remarks were reported on a day when AI-related anxiety has been weighing on investor sentiment. Bessent did not elaborate on specific safeguards or regulatory measures in the account.
TREE NEWS update: Anthropic released a research preview of Claude Duet, a proactive AI assistant powered by Opus 5.5, with a waitlist now open. The assistant runs on its own cloud computer, connects to users’ apps, delivers a morning briefing and keeps working on tasks while users are away, asking for consent before sensitive actions such as sending messages or making payments. The preview is limited to Claude Pro and Max users in the U.S. via iOS, Android and claude.ai; Team and Enterprise plans are excluded.
TREE NEWS update: Walmart will launch an agent for AI shoppers, enabling consumers to shop through apps and agents. The tools will let consumers set spending limits on AI agents. The company did not specify a launch date or further details.
TREE NEWS update: OpenAI said it will bring its frontier models and Codex agentic workflows to Google Cloud’s Model Garden, letting enterprises use its technology within the environments they already build in. The move places OpenAI models on a rival cloud platform. No timeline or model names were specified in the announcement.
TREE NEWS update: Anthropic has signed an agreement with HclTech covering the deployment of its Claude model. The deal makes HclTech the latest enterprise partner to roll out Claude, though the companies did not disclose financial terms or the scope of deployment. The agreement follows a broader push by Anthropic to expand enterprise distribution of its AI models.
TREE NEWS update: OpenAI released a blueprint to empower cyber defenders, building on its earlier cybersecurity action plan and frontier safety blueprint. The company said the new document is based on those two prior frameworks. No further details on the blueprint’s contents or timeline were provided.
TREE NEWS update: Microsoft said that in September 2026 the hacking group Storm-0940 used an artificial intelligence framework called Artex to launch automated attacks against hundreds of organizations worldwide. In several cases the Artex-driven attacks gained privileged cloud access in under three hours. Storm-0940 is assessed to be behind the recently disclosed ConnectWise multi-tenant application supply-chain attack, and its password-spraying campaign targeted 327 accounts at eight organizations between October 10 and 14.
TREE NEWS update: OpenAI said its AI models discovered the PHP2Shell vulnerability and led a two-month coalition to patch it, a flaw serious enough that governments took websites offline. The same models were behind intrusions into Hugging Face and an Australian government portal this year. The disclosure frames OpenAI’s security work as both a defensive tool and a source of risk.
TREE NEWS reports: SpaceX is offering employees a 20% pay premium to help bring its Memphis, Tennessee supercomputer into operation, Business Insider reported. The incentive is aimed at staffing the company’s supercomputing facility in the US city. No further details on the scale of the operation or the number of employees involved were provided.
TREE NEWS reports: Anthropic could publicly file its S-1 registration statement as soon as this Friday, with a listing potentially following in November, subject to market conditions. The company declined to comment on the timing. Wall Street expects Anthropic to be valued at about $2 trillion, with the offering raising up to $100 billion; its draft S-1 showed first-half revenue of $16.3 billion and a net loss of $80.7 billion.
TREE NEWS update: Dataline has raised $10.6 million to build a decentralized data network for AI agents, the startup said. The company also plans to hold a public token sale through Uniswap’s auction platform, with details expected next week. The funding will support the development of the network, which is designed to let AI agents access and share data without a central intermediary.