Search “Google”

OpenAI and Google Cloud reach partnership; frontier models and Codex to launch on Google Cloud

On October 21, 2026, OpenAI announced a partnership with Google Cloud to bring its frontier models and Codex agent workflows to Google Cloud's Model Garden, allowing enterprises to use these technologies within their existing build environments. OpenAI's official statement did not disclose specific launch timing, the list of available models, or pricing information. Subsequent reports said that GPT-6 Astra, GPT-6.1 Sol, GPT-6 Luna, and Codex will launch on Google Cloud, with the first batch of models already available to select enterprise customers in private preview and a general availability plan to roll out in the coming weeks; enterprises can call them via the Google Cloud API and purchase them on Google Cloud Marketplace using their cloud account balances. Other reports said that GPT-6 Astra, GPT-6.1 Sol, and GPT-6 Luna are already live on Google Cloud.

Trump demands data centers generate their own power and pay dividends to communities

On October 21, 2026, Trump posted on Truth Social stating that large new data centers must generate their own power and pay cash "dividends" to their host communities. Related reports noted that the closest existing commitment is the Ratepayer Protection Pledge signed in March of that year by Amazon, Google, Meta, Microsoft, OpenAI, Oracle, and xAI, under which signatories pledged to build or purchase generating capacity and bear infrastructure upgrade costs, though experts believe the pledge may not be enforceable at the federal level. Trump said the dividend proposal goes further than that pledge. Reports on October 22 further stated that Trump also demanded that data centers supply surplus power to local communities free of charge or sell it to them, and noted that communities across the United States have continued to protest AI data center projects over rising electricity prices, noise, water shortages, and other issues. There are currently no facts on subsequent developments.

Apple updates App Store Review Guidelines to address AI non-consensual intimate imagery

On October 16, 2026, Apple updated the App Store Review Guidelines, targeting AI-generated non-consensual intimate imagery and apps that conceal harmful features during review. The revised Guideline 1.2 requires apps hosting user-generated content to take measures to prevent and remove intimate images created without consent, and to provide reporting and blocking channels; Guideline 2.5.19 explicitly states that developers must not misrepresent an app's purpose during review or add undisclosed features after launch, and violations may result in removal of both the app and the developer account. Reports say the adjustment stemmed from investigations finding that such undressing apps were still listed, with cumulative downloads of 483 million and revenue exceeding $122 million for the apps involved, and Apple has removed 15 of them. On October 22, the Tech Transparency Project published an investigation stating that, using four test accounts for 13-year-old boys, it found that even with parental controls enabled, 135 AI apps capable of generating nude images could still be accessed on the Apple and Google app stores; both platforms rely on developer questionnaires to self-report age ratings. As of September 28, Apple had removed 25 apps and Google 6, but all the removed apps remained fully usable on the test phones.

WSJ sees Anthropic's draft IPO filing

The Wall Street Journal has seen Anthropic's draft IPO filing. A report dated October 21, 2026, said the draft showed Anthropic had first-half revenue of $16.3 billion and a book loss of $80.7 billion, with compute commitments of about $450 billion; the draft also disclosed fixed payments totaling $25.2 billion to Google, AWS and Microsoft over the next 7 to 10 years, payable whether or not the capacity is used. A report dated October 22 said the same draft showed first-half revenue of $16.3 billion, up nearly 14-fold year over year, and a net loss of $80.7 billion, of which about $72 billion came from the revaluation of financing instruments such as convertible notes; compute and infrastructure spending in the same period was nearly $9 billion, about 55% of revenue, and forward compute commitments exceeded $600 billion, potentially reaching as high as $750 billion if additional agreements are finalized. There is currently no further progress confirmed by factual evidence.

Andreessen Horowitz leads Eka Robotics Series A funding

On October 20, 2026, Andreessen Horowitz (a16z) led the Series A funding round of Eka Robotics. According to a TechCrunch AI report that day, Eka Robotics completed a $150 million Series A funding round led by a16z and F-Prime, at a post-money valuation of $1.45 billion; the company was founded by MIT professor Pulkit Agrawal and former Google DeepMind researcher Tuomas Haarnoja, and its vision-force-action model, paired with force-sensor grippers, can grasp about 1,500 objects weighing under 1 kg per hour with a success rate exceeding 95%; the grippers cost less than $1,000 and are rated for 5 million cycles, and the company is currently conducting pilot tests of pick-and-place and cable manipulation with industrial customers. The next day, Zhidongxi reported likewise that this funding round was $150 million at a post-money valuation of about $1.45 billion, co-led by a16z and F-Prime with participation from Samsung Next and others, and said the company was founded in 2025 and is carrying out pilot cooperation with multiple industrial customers. There is currently no further progress following the completion of this funding round.

MiniMax thanks for bringing H3 to TPU and looks forward to open-source work on efficient video inference

MiniMax responded to SGLang including MiniMax H3 in the TPU support list of SGLang-TorchTPU, expressing thanks and saying it looks forward to more open-source work around efficient video inference. SGLang said that SGLang-TorchTPU was jointly built by SGLang, the Google TPU team, and the Meta PyTorch team, has been used for Meta's production inference, and supports models including Muse Spark Open, GLM-5.3 Flash, Kimi K3, Ling 3.1 Flash, and MiniMax H3. There are currently no facts about subsequent developments.

Report says Eric Schmidt holds a 1% stake in Anthropic

According to Bloomberg, former Google CEO Eric Schmidt holds about a 1% stake in Anthropic. TechCrunch AI reported on October 21, 2026 that Schmidt participated in Anthropic's Series A funding round in 2021, investing tens of millions of dollars; this stake is about half of the 2% held by co-founder and CEO Dario Amodei. The report also said Anthropic is widely expected to go public before the end of the year, with a maximum valuation expected to reach $2 trillion, at which point the value of Schmidt's stake could increase by as much as $20 billion; neither Schmidt nor Anthropic immediately responded to requests for comment. The next day, Fortune AI reported, citing people familiar with the matter cited by Bloomberg, that if Anthropic goes public at a target valuation of $2 trillion, this stake could be worth $20 billion; the report said Schmidt was introduced to Dario Amodei by Cami Clark and participated in Anthropic's $124 million Series A funding round in 2021, when the company was valued at $461 million. There have been no further developments so far.

Sundar Pichai says Argon usage reached nearly 25 trillion tokens within 72 hours of launch

Sundar Pichai said that within 72 hours of Google's Argon going live, usage approached 25 trillion tokens, roughly 3 times the usage of Gemini 3.0 over the same period. He also mentioned that the model is currently being served, but compute capacity is limited. The remarks came from a video on Alex Kantrowitz's YouTube channel.

Metis Reasoning is founded in London, focusing on a compute-efficient AI approach

On October 20, 2026, the London-based AI company Metis Reasoning was officially established and launched, focusing on a compute-efficient reasoning approach as it seeks a direction distinct from the frontier models of Anthropic and OpenAI. According to Tech.eu, the company was co-founded by Thore Graepel, former head of research at Google DeepMind and co-creator of AlphaGo, and has a team of 12 people spanning researchers, mathematicians, robotics experts and technology entrepreneurs. The company believes that LLMs have fundamental limitations in reasoning, with results that are difficult to verify and lacking an audit trail; it says its approach is more verifiable, more robust and safer, and that it has already achieved some positive results on frontier reasoning problems at a cost far below that of frontier models. The report says its initial funding is believed to be in the tens of millions of dollars, but Metis has not yet announced a release date for its first AI model.

Xbow uses AI to discover Chrome vulnerability chain, wins Google bounty

Xbow USA researchers spent six days developing a Chrome exploit chain that first escapes the sandbox hosting the V8 JavaScript engine, then escapes Chrome's operating system process sandbox, and can launch Windows applications from within a malicious webpage; Google has already fixed the related flaws. They chose OpenAI's GPT-5.6 Sol to search the Chrome source code, validate ideas, and build a proof of concept, while strategy formulation, search direction, and verification of results remained the responsibility of the researchers. The AI compute cost $40,000, and they received a $250,000 Google bug bounty, the first under the new 2026 rules.

OpenCode discloses distillation attack affecting 1.4% of console users

OpenCode disclosed that a security incident related to a large-scale distillation attack occurred the previous night, affecting 1.4% of OpenCode Console users—specifically those who had linked external service providers. The company notified all affected users by email on the morning of the 18th; those who did not receive an email are not affected. The incident is still being handled, and OpenCode said it is working with Anthropic, OpenAI, and Google to remediate it. The same day, Nathan Lambert raised doubts, arguing that it is strange to lump a cyberattack that stole API keys together with a "model transfer station" under the name "distillation," and that it is unclear how this relates to model training. Subsequently, Anthropic's Lydia Hallie said the team is working with OpenCode to handle the leak incident, has revoked the Claude API keys it believes were affected, and recommends that users who previously saved keys in OpenCode create new keys.

English translation is being prepared.

Chollet says Argon performs similarly to Astra on ARC-AGI-3

François Chollet commented that Gemini 4 Argon performs similarly to Astra on ARC-AGI-3, with Google returning to the frontier. On the same day, ARC Prize published the verified results for Gemini 4 Argon on the ARC-AGI series: ARC-AGI-3 scored 97.5% under the provider adapter harness at a cost of $4.8K; under the standard harness it scored 49.8% at a cost of $5.8K. ARC-AGI-2 scored 91.3% at $0.32 per task; ARC-AGI-1 scored 98.5% at $0.16 per task. ARC Prize said Google DeepMind is the second frontier AI lab to exceed 95% on ARC-AGI-3, and Argon's 97.5% is second only to GPT-6 Astra's 99.9%. The standard harness test could not complete all 55 games due to API issues; the high-reasoning setting completed only 19 games, so no score was published.

English translation is being prepared.

François Chollet commented that Gemini 4 Argon performs similarly to Astra on ARC-AGI-3, with Google returning to the frontier. On the same day, ARC Prize published the verified results for Gemini 4 Argon on the ARC-AGI series: ARC-AGI-3 scored 97.5% under the provider adapter harness at a cost of $4.8K; under the standard harness it scored 49.8% at a cost of $5.8K. ARC-AGI-2 scored 91.3% at $0.32 per task; ARC-AGI-1 scored 98.5% at $0.16 per task. ARC Prize said Google DeepMind is the second frontier AI lab to exceed 95% on ARC-AGI-3, and Argon's 97.5% is second only to GPT-6 Astra's 99.9%. The standard harness test could not complete all 55 games due to API issues; the high-reasoning setting completed only 19 games, so no score was published.

View story:Chollet says Argon performs similarly to Astra on ARC-AGI-3

English translation is being prepared.

English translation is being prepared.

Gemini 4 Argon appears on the GCP quota page

On October 13, 2026, Gemini 4 Argon appeared on the GCP quota page. Testing Catalog, citing AiBattle, said the model was expected to launch within a few hours and called it a precursor to an imminent API model release; earlier that day, a Google employee had also teased an upcoming release. That evening, Testing Catalog reported that the model had begun rolling out and was already live in GCP's Agent Studio, where users could test it, with Google AI Studio and the Gemini app expected to follow. On October 14, Google officially released Gemini 4 Argon, open to developers and enterprises worldwide, usable via AI Studio and the Gemini platform API, and supported in Google Antigravity as well as the Gemini app and Search AI Mode for Ultra and Pro subscribers; reports said it supports long decoding output of up to 1,048,576 tokens and announced limited-time introductory pricing.

English translation is being prepared.

English translation is being prepared.

Anthropic releases Claude Opus 5.5

On September 22, 2026, Anthropic released Claude Opus 5.5. According to official materials, the model ranks first on Artificial Analysis's Intelligence Index v4.3 with a score of 58, 7 points higher than Claude Opus 5, and tops Vals AI's Vals Index at 69.69%; it supports up to 1 million tokens of input and 128,000 tokens of output, with reasoning enabled by default across five levels, and API pricing at $4/$0.25/$20 per million input/cached/output tokens. Earlier reports said its output price was $20 per million tokens, lower than the previous generation's $25. On September 23, Yuchen Jin said it was about 30% faster and about 40% cheaper per task than Opus 5. On September 25, Thomas Wolf said he used Opus 5.5 to generate a video from a Google Doc. On October 4, Ado Kukic said that using Opus 5.5 at medium and high effort he did not hit the 5-hour limit, using at most about 74% in a single session. On October 9, Ado again said that the weekend after Opus 5.5's release was his most productive coding day of the year so far. During the same period, Anthropic also released Sonnet 5.5 and Haiku 5.5, and lowered Sonn