NAll news

All news

Sunday45 items

  1. Phoenix Technology
    AI score50

    Anthropic Bans Claude from Recreating Existing Games in Opus 5.5 System Prompt

    Anthropic confirmed that it added two sentences to the Opus 5.5 system prompt instructing Claude not to recreate specific games belonging to others. This came after blogger chasm discovered that Claude Code was blocking game clones. The rule only constrains Claude itself; users can switch to other models or break tasks down into specific technical tasks whose context does not reveal the overall goal to bypass it. After the ban was announced, a large number of original developers complained on the Claude Code GitHub issue page, saying they were refused when developing original games on the grounds that they were suspected of cloning existing games.

    Same story · 15 reports

    #Safety / alignment#English text is being prepared.#Agents

  2. Sina Tech
    AI score62

    Wall Street AI hedging sentiment heats up, demand for cloud providers' CDS and put options surges

    Some Wall Street investors are hedging against the risk of an AI rally reversal through put options and credit default swaps, with the cost of debt default protection for cloud providers such as Amazon, Alphabet, Microsoft, Meta and Oracle rising to its highest level since 2021. The scale of CoreWeave-linked options contracts hit a record high this month, with the most actively traded being put options expiring in 2028 with a strike price of $35. Jump Trading has newly established a compute trading desk, designing over-the-counter derivatives for clients to hedge against the risk of falling compute prices.

    #Industry news#English text is being prepared.#OpenAI

  3. 36Kr · AI
    AI score81

    OpenAI's new funding target cut from $30 billion to $20 billion

    According to multiple media reports, OpenAI has cut its new funding target from $30 billion to $20 billion, with its valuation dropping from $1.4 trillion to $1.25 trillion. As of Friday's commitment deadline, actual investment commitments stood at about $11 billion, less than half of the original target. This round is led by a UAE consortium, with UAE state-owned technology funds such as MGX as core investors, and Founders Fund, Andreessen Horowitz, SoftBank and others participating; OpenAI plans to expand fundraising to family offices and high-net-worth individual investors.

    Same story · 8 reports

    #Industry news#OpenAI#Anthropic

  4. ServeTheHome
    AI score62

    NVIDIA Adjusts Rubin GPU Specs at Mass Production: HBM4 Bandwidth Reduced to 19.2TB/s…

    When the Rubin GPU entered mass production, NVIDIA lowered its official specs: per-card HBM4 memory bandwidth dropped from 22TB/s to 19.2TB/s (about 11%), NVLink bandwidth dropped from 3.6TB/s to 3.0TB/s (about 17%), while scale-out network bandwidth increased from 0.4TB/s to 0.45TB/s. The Vera Rubin NVL72 system was adjusted in tandem: memory bandwidth dropped from 1580TB/s to 1400TB/s, NVLink dropped from 260TB/s to 216TB/s, network bandwidth rose from 28.8TB/s to 32.4TB/s, and HBM4 capacity remained at 20.7TB. The HGX Rubin NVL8 was hit hardest: HBM4 capacity was cut from 2.3TB to 2.0TB, memory bandwidth fell by about 26%, and FP4 training performance dropped from 280 PFLOPS to 270 PFLOPS.

    #Product updates#English text is being prepared.#English text is being prepared.

  5. The Paper
    AI score56

    SiliconFlow's Hong Kong IPO filing accepted, aims to become the first AI Token factory stock

    The Paper learned from the website of the China Securities Regulatory Commission that the Hong Kong stock issuance and listing filing status of Beijing SiliconFlow Technology Co., Ltd. is "accepted," with a material receipt date of October 19, 2026. The company submitted a listing application to the Main Board of the Hong Kong Stock Exchange on June 30 this year, intending to list under Chapter 18C, with Huatai International and Guotai Haitong serving as joint sponsors. The prospectus shows that SiliconFlow's revenue in 2024 and 2025 was 7.346 million yuan and 55.33 million yuan, respectively, with losses for the year of 81.915 million yuan and 345 million yuan, respectively. At the end of 2025, total liabilities were 693.7 million yuan, of which redeemable liabilities were 498.6 million yuan.

    #Industry news#English text is being prepared.#DeepSeek

  6. Phoenix Technology
    AI score62

    US Department of Justice investigates case of AI fake students defrauding $200 million in student loans

    The US Department of Justice announced the results of "Operation Ghostbusters," an investigation into student loan fraud that used stolen identities and artificial intelligence to complete applications and pose as students enrolled in online courses, with the US government estimated to have lost more than $200 million as a result. Multiple community colleges and universities have suspected fraud in the Free Application for Federal Student Aid (FAFSA) program since 2021; five people have been charged so far, and more than 40 open cases involve 26 FBI field offices and 20 US Attorney's Offices.

    Same story · 3 reports

    #Industry news#Safety / alignment#English text is being prepared.

  7. Phoenix Technology
    AI score71

    OpenAI's funding target shrinks from $30 billion; Anthropic reportedly unlikely to obtain investment-grade rating after IPO

    According to The Wall Street Journal, in OpenAI's new funding round, which originally targeted $30 billion, investors were willing to provide only $11 billion, cutting the funding target by one-third; Sam Altman said he will continue reaching out to family offices and high-net-worth individuals in the coming weeks. Blackstone, a major investor in fellow AI company Anthropic, changed its tune and said that Anthropic will likely be unable to obtain an investment-grade rating after going public, which would affect its arrangements to borrow at low cost for its plan to build more than $800 billion in computing infrastructure by 2045. On the day the news emerged, the Nasdaq 100 index closed down 3.5%, bringing its cumulative decline for the entire month of October to about 5.3%.

    #Industry news#OpenAI#Anthropic

  8. IT Home · Artificial Intelligence
    AI score55

    White House Science and Technology Advisor Council Chairman Sachs Comments on Anthropic S-1: May Become the Greatest Company in the History of Capitalism

    David Sachs, Chairman of the U.S. White House Science and Technology Advisory Council, analyzed Anthropic's S-1 registration filing on the All-In podcast, saying that this artificial intelligence company is likely to become the greatest company in the history of capitalism. The filing shows that Anthropic had revenue of $16.3 billion in the first half of 2026 and cash on its books of $112 billion, with projected 2026 annual recurring revenue reaching $100 billion, and seven founders holding 50.1% of voting rights through Class F shares. Sachs believes that its abundant cash makes its IPO financing needs lower than outside expectations, thereby giving it stronger pricing power.

    #English text is being prepared.#Industry news#Anthropic

  9. English text is being prepared.
    AI score57

    Report: Spotify plans to launch AI music remix service early next year, allowing users to create their own remixes and covers

    According to the Financial Times, Spotify plans to launch an AI-driven music remix service in early 2027, allowing paid subscribers to create their own remixes and cover versions of songs, and is considering charging an additional $4 to $5 per month. Spotify is negotiating licensing agreements with Sony Music and Warner Music, has reached agreements with Universal Music Group, Merlin, and Believe, and artists themselves can still decide whether to participate. Under the proposed plan, royalties generated from remixes will flow in full back to the original artists, and plays of remixed works will also count toward the original songs' chart performance.

    Same story · 2 reports

    #Industry news#Speech#English text is being prepared.

  10. IT Home · Artificial Intelligence
    AI score60

    US Department of Justice Investigates AI Fake Students Defrauding $200 Million in Student Loans

    The US Department of Justice announced the results of "Operation Ghostbusters," an investigation into student loan fraud that used stolen identities and artificial intelligence to complete applications and pose as students enrolled in online courses. Since 2021, multiple community colleges and universities have suspected fraud in the federal student aid (FAFSA) program, with estimated losses to the US government exceeding $200 million. Five people have been charged so far, and more than 40 open cases involve 26 FBI field offices and 20 US Attorney's Offices.

    Same story · 3 reports

    #Industry news#English text is being prepared.

  11. English text is being prepared.
    AI score56

    Over 70,000 AI games on the Roblox platform; more than 80% of the top 100 titles use AI-assisted development

    Roblox revealed in its fiscal year 2026 third-quarter report that the total number of AI games created using Build on the platform has exceeded 70,000, and more than 80% of the games ranked in the top 100 by playtime in September used AI-assisted development. More than 500,000 users have already used the Build tool, and this quarter two games created almost entirely by AI entered the platform's top 20 ranking by concurrent online players. Build was launched at the end of July 2026, features a mobile-first design, is open to users aged 9 and above who have completed age verification, and published AI-generated games are only available to users aged 16 and above.

    #English text is being prepared.#English text is being prepared.#Industry news

  12. Sina Tech
    AI score63

    OpenAI's funding target shrinks, Anthropic reportedly faces rating setback, Nasdaq falls 3.46% in a single day

    According to The Wall Street Journal, OpenAI's new funding round target has been lowered from $30 billion, with investors willing to provide only $11 billion, cutting the funding target by about one-third. Separately, according to the Financial Times, Anthropic's major investor Blackstone changed its tune, saying the company may not obtain an investment-grade rating after going public; on the day of the news, the Nasdaq 100 index closed down 3.5%, and the Nasdaq fell about 5.3% cumulatively over the entire month of October. The article also mentions that Amazon's trailing 12-month free cash flow was -$42.4 billion, and compares this round of adjustment to the internet bubble 26 years ago.

    #Industry news#OpenAI#Anthropic

  13. Phoenix Technology
    AI score60

    OpenAI Announces ChatGPT and Codex Will Share Quota Starting in December

    OpenAI announced that starting in December this year, the quotas for its three paid plans—ChatGPT Plus, Pro, and Business—will be calculated as a combined total, with ChatGPT and Codex sharing the same quota pool and billed uniformly by token; consumption depends on the model selected and task complexity. Instant chat does not consume the shared quota, and users can check their remaining quota and reset time on the "Settings—Usage" page in ChatGPT. This adjustment is related to OpenAI integrating ChatGPT Work's agent capabilities into ChatGPT, and the company expects most users will not run out of quota more often than they do now.

    Same story · 13 reports

    #Product updates#OpenAI#Industry news

  14. Phoenix Technology
    AI score76

    Blackstone expects Anthropic to struggle to obtain an investment-grade rating after going public, potentially falling into the BB junk-grade category

    Blackstone told investors that Anthropic may not immediately obtain an investment-grade rating after going public, and that major rating agencies are more likely to assign a high-end junk rating in the BB range. If the rating falls into junk territory, Anthropic's prospects for financing its more than $800 billion computing power plan will become more complicated, because funding pools such as insurance companies and pension funds will narrow significantly and interest rates will be higher. In the $63 billion financing codenamed "North Star," more than $42 billion of senior loans are guaranteed by Broadcom, while $18 billion of subordinated debt is backed by Anthropic's lease payments; KBRA assigned a BB+ rating and Morningstar DBRS assigned a BB rating.

    #Industry news#OpenAI#Anthropic

  15. Phoenix Technology
    AI score68

    Blackstone Reverses Stance, Says Anthropic May Struggle to Obtain Investment-Grade Credit Rating After IPO

    According to a report by the UK's Financial Times, Blackstone Group has told investors that Anthropic is more likely to receive a speculative-grade rating in the BB range after its IPO, rather than the investment-grade rating previously expected. This reversal emerged in Anthropic's $63 billion chip financing deal; a speculative-grade rating would restrict investment by institutions such as insurance companies and pension funds. The prospectus shows that Anthropic plans to invest more than $800 billion in computing infrastructure by 2045, and currently holds more than $112 billion in cash and cash equivalents.

    Same story · 4 reports

    #Industry news#Anthropic

  16. Phoenix Technology
    AI score68

    Microsoft, Google, Oracle and Two Other Tech Giants Commit Over $3 Trillion to AI Infrastructure

    According to Bloomberg, five major tech giants—Alphabet, Meta, Microsoft, Amazon and Oracle—have committed to investing more than $3 trillion in data centers and other areas over the coming years. Among them, Alphabet's total contract commitments exceed $1 trillion, nearly eight times the level a year earlier; Meta's total future spending exceeds $800 billion, nearly half of which is for data center leases that have not yet taken effect and are payable over terms of up to 30 years. The five companies' combined future lease payment commitments total $1.3 trillion, and other purchase commitments total $1.9 trillion. Oracle has already seen negative free cash flow due to its investment in AI data centers.

    Same story · 3 reports

    #Industry news#English text is being prepared.#OpenAI

  17. East Money
    AI score56

    S&P Downgrades Meta's Rating Outlook to Negative on AI Spending Pressure

    S&P Global Ratings downgraded the outlook on Meta's AA- credit rating from stable to negative on October 31, citing continued increases in AI-related spending and rising leverage. S&P expects Meta's capital expenditure to reach $140 billion this year and exceed $225 billion in 2027, with leverage rising above 1x as early as next year, and said Meta's ability to convert AI investment into commercial revenue is relatively limited. A Meta spokesperson said the company disagrees with this assessment and believes future product and business opportunities are sufficient to support its credit profile.

    #Industry news#OpenAI#Meta

  18. East Money
    AI score74

    OpenAI's target valuation lowered to $1.25 trillion, funding target cut to $20 billion

    OpenAI has lowered the target valuation for this funding round from $1.4 trillion to $1.25 trillion, a reduction of $150 billion, or about 10.7%; the funding target has also been cut from $30 billion to $20 billion, a reduction of $10 billion. People familiar with the matter said that because some investors reduced their subscription amounts, the actual funds raised in the first closing are expected to be less than $20 billion. MGX, part of a consortium of UAE investment institutions, is expected to become the largest participant, while Founders Fund, Coatue, Sequoia Capital, Andreessen Horowitz and SoftBank also plan to participate.

    Same story · 8 reports

    #Industry news#OpenAI

  19. Simon Willison
    AI score65

    Parameter Labs discloses an exploit chain that can take over ChatGPT accounts

    Parameter Labs disclosed an attack chain targeting the ChatGPT HTML preview (formerly Canvas) sandbox. An attacker can craft a share link that, when visited by another logged-in user, steals private data or performs actions on their behalf. The attack starts with the postMessage-receiving code of the MathLive virtual keyboard using innerHTML, injects an iframe without sandbox restrictions, and then uses the YouTube oEmbed endpoint allowed by CSP to reflect and execute JavaScript. OpenAI has fixed the vulnerability and paid a $5,000 bounty.

    #Safety / alignment#OpenAI

  20. Rohan Paul
    AI score71

    OpenAI will merge ChatGPT and Codex quotas into a single shared token allowance in December

    OpenAI announced on its official documentation page that ChatGPT and Codex under the Plus, Pro, and Business plans will share a single token allowance, with the change starting in December. At that time, OpenAI will merge the agent features of ChatGPT Work, an agent built on Codex in July, into the main ChatGPT experience. Currently, Work already shares limits with Codex, while regular ChatGPT conversations are metered separately, so heavy coding does not consume chat quota; after the switch, each request will be priced by token, whereas the current limits for regular conversations are calculated by the weekly message count for each model, such as the weekly cap for GPT-6 Pro messages.

    Same story · 13 reports

    #Product updates#OpenAI#Agents