Flash News

TREE NEWS update: World released a World ID suite aimed at banks and made its World ID for Zoom feature generally available after a beta that began in April. The Zoom tool is designed to spot deepfaked participants on video calls. The bank-focused suite extends World ID verification to financial institutions.

TREE NEWS update: Identity network World is packaging its video-call deepfake detection, identity wallet and AI-agent tools for financial institutions, the company said, with four design partners already exploring uses. The bundle is aimed at banks seeking to verify that customers are human as AI-generated impersonation grows. World did not disclose pricing or a launch date for the offering.

TREE NEWS update: Bank of Japan Governor Kazuo Ueda said he expects AI-related demand to trigger demand-pull price increases. Ueda’s remark points to the central bank watching AI-driven demand as a potential source of inflationary pressure. No further details or policy implications were provided in the material.

TREE NEWS reports: DWF Labs said the market has not fully priced in the potential of AI and crypto convergence, noting that frontier capital is still flowing mainly into space, robotics and AI rather than crypto. The firm said agent trading and payments are picking up, and that compute markets running on crypto networks could become a theme over the next year as AI capital expenditure and model development move into execution.

TREE NEWS update: Situational Awareness, the fund run by “AI stock god” Leopold, will disclose its third-quarter equity holdings on Nov 16. The fund reportedly returned 439% in the first half on leveraged bets on SanDisk (SNDK) and Micron (MU), but a July selloff in AI stocks cut its portfolio by about 67% in a single month, after which it sold most of its public equity positions, with Citadel taking over a substantial portion.

TREE NEWS reports: Anthropic co-founder Jack Clark warned that artificial intelligence could raise US GDP from about $30 trillion to $44 trillion, but humanity may have only a few years to establish effective safeguards to keep AI from exceeding human control. Clark also cautioned that AI could trigger a so-called singularity and called for global rules to prevent misuse in biological research and cyberweapon development. Anthropic is reportedly planning a November IPO at a valuation that could reach $2 trillion.

TREE NEWS update: Anthropic has signed a 20-year data center lease with Canada’s EverWind Fuels to support the operation of its Claude AI models. The deal, Anthropic’s first data center agreement in Canada, covers a 95-megawatt first-phase facility at the Point Tupper campus in Cape Breton, Nova Scotia; the lease value was not disclosed. The first phase is slated to come online in late 2027, with a 325-megawatt expansion planned for 2028, bringing the campus to a planned 420 megawatts subject to commercial, technical and regulatory conditions.

TREE NEWS update: DeepSeek said its API now supports web search and fetch, billed at 5 yuan per 1,000 successful requests. Search and fetch calls made through DeepSeek’s server-side tools in Harness, Claude Code and Codex are currently free, but will be charged per call from 00:00 Beijing time on Nov. 1, 2026.

TREE NEWS update: DeepSeek said on Oct 18 that its API now supports web search and page fetching, billed at 5 yuan per 1,000 successful requests. Search and Fetch calls made through Server-side tool use in DeepSeek Harness, Claude Code and Codex are currently free but will be charged per request starting Nov 1 at 00:00.

TREE NEWS reports: Samsung Electronics will use TSMC’s 3-nanometer process to produce base dies for Nvidia’s next-generation custom HBM chips, marking Samsung’s first collaboration with TSMC at the HBM manufacturing stage. Nvidia has asked Samsung, SK Hynix and Micron Technology to manufacture NVHBM base dies using TSMC’s process and to produce their core chips using sixth-generation DRAM technology.

TREE NEWS reports: Microsoft has withdrawn from a data centre partnership with UK cloud provider Nscale in Lawton, Essex, a setback for the startup as it prepares an IPO. Microsoft gave no reason for ending the deal, which had faced delays; local authorities told Nscale the facility may not get power until the 2030s. Nscale’s flagship Norway project is about six months behind schedule, with only 6.4 MW of 30 MW running. Its IPO filing shows about 85% of contracts depend on Microsoft and Anthropic.

TREE NEWS update: OpenAI is developing humanoid robots, with CEO Sam Altman saying the company aims to produce an impressive demonstration prototype by 2027. Altman said robot intelligence systems are the project’s core focus, but real-world public use is still years away. The project remains far from a commercial product, and details on the demo, robot design and public access timing have not been disclosed.

TREE NEWS reports: Microsoft has cancelled a deal with Nscale covering data centers in the UK. The report did not disclose the value of the agreement or the reason for the cancellation. Nscale is a UK-based AI infrastructure provider that builds and operates data centers for artificial intelligence workloads.

TREE NEWS update: Tesla CEO Elon Musk acknowledged he went “somewhat off track” in political affairs and spent more time away from his companies than he should have. Musk said he should have played a less high-profile role and devoted no more than one or two months to government work, but instead spent about four months at the White House. He added that he should have returned sooner to focus on artificial intelligence.

TREE NEWS reports: SoftBank is considering an acquisition of Allied Universal, the largest US security-services company, at a potential equity valuation of up to about $18 billion, with roughly $12 billion in publicly traded loans and bonds bringing total transaction value close to $30 billion. Masayoshi Son is exploring AI robotics applications across industries, and SoftBank hopes robots could eventually take over some of Allied’s security duties, though humanoid security robots may take longer to deploy. Discussions are still early and terms, including price, are not final.

TREE NEWS reports: Nvidia, OpenAI, Anthropic, Google, Meta and xAI signed a White House frontier AI agreement on Sept. 29, but it lets each company choose its own auditor and set the scope of review, Samuel Vance-Law, research director at the Decentralization Research Center, said in a commentary. The pact does not specify who selects auditors, what can be examined or whether findings are published, and carries no penalties for breach.

TREE NEWS update: StepFun opened pre-orders in China on Oct. 13 for its STEPX Neo smartphone, priced at 3,999 yuan. The phone runs an AI agent built directly into the operating system rather than offered as a separate app. The launch marks the Chinese startup’s move into AI-agent-first handset hardware.

TREE NEWS reports: The Rollup founder Andy said on X that an AI product designed for institutions moving on-chain and meeting specific risk-management policy requirements could launch within days, with several funds and asset managers already participating in trials. No launch date, product name or participating institutions were disclosed.

TREE NEWS update: Chinese Academy of Sciences academician and Peking University chair professor E Weinan said at the Guanghua Thought Leadership Forum that AI, essentially ‘mathematics plus computer engineering,’ will sharply lower research barriers and drive ‘research democratization,’ allowing any capable team or individual to participate in innovation through open labs and cloud services. He argued that insufficient innovation capacity is the root cause of corporate involution and that an AI-native industrial software system could match or surpass Siemens within one to two years.

TREE NEWS update: KBRA assigned Anthropic a BB+ issuer rating in June, the highest speculative-grade level and one notch below investment grade, with a stable outlook. The rating covers about $4.4 billion of subordinated debt in a $35 billion chip financing package, yielding roughly 8.5% versus 5.75% for senior debt. KBRA substituted a risk score for a full financial model because Anthropic disclosed limited data. The five-year-old company had zero debt on its balance sheet at the end of June.