On mvanemmerik.com, Nov 16: NFL Week 10 chalk talk: Detroit dazzles, Cincinnati takes the North
vanemmerik.aiClaude, ChatGPT and Grok news, twice a day

Wednesday, November 11, 2026

xAI releases Grok Bot 1.0, taking its always-on agent out of beta with shared team bots

xAI's always-on agent left beta, adding shared bots that work in Slack, Teams and meetings. OpenAI reportedly told investors Q3 revenue hit about $10 billion while its operating loss widened. Anthropic's bankers are reportedly weighing a higher IPO range, the company opened the offer to Australian retail buyers, kept a now-cancellable Lambda compute deal and gave paid Claude users a free usage reset. Researchers also found thousands of Grok Build apps open to anyone, and Grok API watermarking is now off by default.

Wednesday, November 11, 2026

GrokLaunch

xAI releases Grok Bot 1.0, ending the beta and letting teams share bots in Slack, Microsoft Teams and video calls

xAI posted Grok Bot 1.0 on Nov. 11, nearly three months after the agent entered beta. Bots can now be shared across a team, work inside Slack, Microsoft Teams and video calls, offer to take on work they spot and follow up on unfinished tasks, and xAI says they are faster and more reliable in the browser and use fewer resources. Individuals can create their own bot; Team Bots are on Team and Enterprise plans, and meeting attendance is on Team, Enterprise and Ultra, with Enterprise admins switching it on. The announcement gives no new prices and does not name the model behind it; its page carries a Nov. 2 date, though the launch post went out Nov. 11.

Why it matters: It moves xAI's always-on agent out of beta and into shared team use, the same ground Anthropic and OpenAI are working on with Claude Duet and ChatGPT Dots.

Try it: Update the Grok Bot app or open Grok Bot on the web; on a Team or Enterprise plan, share a bot with colleagues and add it to Slack or Teams.

Grok BotGrokxAIHappened Nov 11, 2026 · Sources: xAI: Introducing Grok Bot 1.0 · CellCog: Grok Bot 1.0, what changed since the beta

ClaudeBusiness

Anthropic's bankers reportedly weigh raising the $70 to $78 IPO price range on strong demandReported

The Financial Times reported, citing five people familiar with the matter, that underwriters led by Goldman Sachs and Morgan Stanley are considering lifting Anthropic's IPO price range above the $70 to $78 a share set on Nov. 10, which could push the valuation target to $1.5 trillion or more. TokenPost separately said prospective buyers have discussed valuations of $1.2 trillion to $1.3 trillion. Anthropic has not commented, and no new range has been filed.

Why it matters: A higher range would raise how much Anthropic collects in what is already set to be one of the largest IPOs on record.

AnthropicHappened Nov 11, 2026 · Sources: Seeking Alpha, citing the Financial Times · TokenPost

ClaudeFeature

Anthropic gives Pro, Max, Ultra and Team users one free full usage reset, good until Dec. 10, after fixing over-counting bugs

Anthropic's ClaudeDevs account said late on Nov. 10 that every Pro, Max, Ultra and Team subscriber gets one free reset of all usage limits across all models, to be used before Dec. 10 at 2 p.m. Pacific. The company said it had fixed a few small issues that used up more of people's limits than intended, without naming them or saying how many users were hit.

Why it matters: It compensates subscribers for limits drained by bugs, a recurring complaint across all three labs' paid plans.

Try it: Look for the reset under Settings > Usage in Claude and use it when you're close to a limit.

ClaudeClaude CodeHappened Nov 10, 2026 · Sources: Startup Fortune

ClaudeBusiness

Anthropic opens its IPO to Australian retail investors through CommSec and Macquarie until Nov. 17Reported

Anthropic lodged an offer document with Australia's securities regulator and opened a retail offer through brokers led by CommSec, alongside Macquarie, on Nov. 11 Sydney time, with applications expected to close Nov. 17, Bloomberg reported. Shares are marketed at $70 to $78, with the final price expected around Nov. 19 and Nasdaq trading as ANT around Nov. 20; there is no ASX listing. Seeking Alpha reported that strong demand has bankers weighing a higher range.

Why it matters: It is a rare chance for non-US individuals to buy into a US tech IPO at the offer price, and a signal of how much retail demand the listing is drawing.

AnthropicHappened Nov 11, 2026 · Sources: Bloomberg via Yahoo Finance · Seeking Alpha

OpenAIBusiness

OpenAI reportedly told investors its Q3 revenue rose about 50% to $10 billion while its operating loss widened to $10.9 billionReported

OpenAI's third-quarter revenue reached roughly $10 billion, up nearly 50% from about $6.7 billion in the second quarter, according to figures the company shared with investors and reported on Nov. 10 and 11 by outlets including DIGITIMES and Dataconomy. Its operating loss grew to about $10.9 billion from $10.1 billion, though its adjusted operating margin improved to about -109% from -150%; Dataconomy adds that enterprise revenue doubled from the prior quarter. OpenAI has not published the figures itself.

Why it matters: The numbers land as Anthropic markets its IPO with a filed Q3 revenue of about $17 billion, making the two companies' finances easy to compare for the first time.

OpenAIHappened Nov 10, 2026 · Sources: DIGITIMES · Dataconomy

GrokSafety

Researchers find 35,578 Grok Build apps open without sign-in, some exposing SpaceXSI sales plans and medical recordsReported

Security firm Claroc said on Nov. 10 that 35,578 of the apps people published from Grok's Build mode at grok.me addresses, 94.7% of those it reached, opened without any login, because the publish panel it tested preselected "Anyone with the link" and app names could be found in passive DNS data. It estimates about one in ten open apps held sensitive material, including 24 built by SpaceXSI and X staff, a hospital staff roster and an ICU patient's lab results. SpaceXSI told a reporter apps stay private until shared and took down the 24 staff apps, but per Claroc had not removed other users' apps or committed to notifying their owners.

Why it matters: App builders that publish with one click can leak real company and patient data when sharing defaults are open.

Try it: If you've published a Grok Build app, check its access setting and switch it to "Only you" unless it is meant to be public.

GrokxAIHappened Nov 10, 2026 · Sources: Claroc Research: GhostShares 2, Grok Build apps exposed

ClaudeBusiness

Lambda misses a Texas lease deadline, letting either side cancel $34.9 billion of its Anthropic compute deal; Anthropic says it will stayReported

Lambda did not secure a data center lease by the deadline in its roughly $41.3 billion August agreement with Anthropic, so either party can now end the $34.9 billion GPU cloud portion, which was meant to supply about 350 MW of Nvidia capacity at Hut 8's Beacon Point site in Texas, TokenPost reported. Anthropic said on Nov. 10 that it does not plan to terminate the deal. The lease condition was already disclosed in Anthropic's IPO filings.

Why it matters: Compute deals this large are central to Anthropic's IPO story, and this one is now cancellable at will.

AnthropicHappened Nov 10, 2026 · Sources: TokenPost · Cape Fear Advisors: Anthropic S-1/A analysis

ClaudePolicy

Anthropic signs enforceable privacy and safety terms with the AFT for school districts using Claude for Teachers

Anthropic announced on Nov. 10 contract terms, negotiated with the American Federation of Teachers and the United Federation of Teachers, for districts using Claude for Teachers: no training on student or educator data except some research and safety uses, a lasting duty to report and contain breaches, data minimization with extra review for features like memory, and human review of AI-informed grading or discipline decisions, Education Week reported. Districts must opt in for future licenses, and the terms don't cover other companies' tools built on Claude. They mirror terms Microsoft signed in September; the AFT says it is still negotiating with OpenAI.

Why it matters: Union-backed contract terms give districts something enforceable, rather than a privacy policy that can change.

ClaudeAnthropicHappened Nov 10, 2026 · Sources: Education Week

GrokPolicy

Update: xAI's docs now say Grok API text watermarking is off by default, with team admins able to switch it on per model

xAI's watermarking page, last updated Nov. 9, now says output watermarking is off by default for grok-4.7, grok-4.6, grok-4.5, grok-4.3 and three grok-4.20 variants, reversing the default-on setting it described for Nov. 7. Team admins can turn it on or off per model in the SpaceXSI Console, changes apply to new requests within five minutes, and some models may be locked on. The page does not explain the change.

Why it matters: Developers who expected watermarked Grok text by default now have to opt in, ahead of EU AI Act marking rules in December.

xAI APIGrokHappened Nov 9, 2026 · Sources: xAI Docs: Output Watermarking