Scoop
AI tool news · rumor vs. reality

The AI Wire ●

Rumors tracked. Announcements verified. Updated daily.

2026-11-04

ANNOUNCED

Meta acquires Conductor, the startup that runs teams of AI coding agents in parallel

Meta has acquired Conductor, the Y Combinator startup whose app lets developers run several AI coding agents at once, each in its own isolated copy of a codebase. Axios reported the deal on November 4 and Meta confirmed it to TechCrunch the same day; financial terms were not disclosed.

Conductor ships no model of its own. It orchestrates first-party agents — Claude Code, Codex, Cursor and OpenCode — through one interface, locally on a Mac or in cloud workspaces that can keep working for hours. The company raised a $22 million Series A in March 2026 and says more than 100,000 people use it.

CEO Charlie Holtz and CTO Jackson de Campos join Meta Superintelligence Labs, working with Nat Friedman and Alexandr Wang on what Holtz calls "the next interface" for AI coding, promised for early next year. The current product winds down fast: new sign-ups stopped November 4, Conductor Cloud goes read-only November 13 and shuts down November 20, and the local app loses support in early 2027.

RELEASED

Google brings Gemini and AI Mode direct checkout to Canada, starting with Lululemon, Shoppers Drug Mart and Simons

Google launched direct checkout in Canada on November 4, letting shoppers buy products inside the Gemini app and AI Mode in Search without leaving for a separate store website. The first retailers on board are Lululemon, Shoppers Drug Mart, Simons and select Shopify merchants, following earlier rollouts in the United States and Australia.

Payments run through Google Pay using cards and shipping details already saved in Google Wallet, and retailers remain the seller of record. The feature is built on the Universal Commerce Protocol, the open standard Google launched with industry partners earlier this year to give shopping agents, retailers and payment providers a common language.

At launch the Canadian tool is limited to one item at a time, with multi-item carts and price alerts planned to match the US feature set. Google says 59 per cent of Canadian holiday shoppers plan to use an AI tool to help them shop this season, and that over the coming months it will work with retailers on extras such as applying loyalty rewards at checkout.

RUMOR Rumor still unresolved

Rumor: ChatGPT's checkout code lists a 'Family Basic' plan lineup from $8 to $500 — OpenAI has announced nothing

ChatGPT's web checkout pricing config now carries a 'Family Basic' lineup, according to Tibor Blaho, who tracks changes in OpenAI's app code and posted a screenshot of the config on November 4. The config shows five entries, from family_basic_go to family_basic_promax, priced at $8, $20, $100, $200 and $500 a month, tax exclusive.

Each amount exactly matches an individual plan OpenAI already sells — Go, Plus, Pro 100, Pro 200 and Pro 500 — and the leak includes no seat count, no feature list and no customer-facing name. That leaves the central question unanswered: a family plan normally implies several people covered, and nothing in the screenshot says how many.

OpenAI has announced nothing, and its pricing page still says Go, Free and Plus are designed for individuals, with sharing sold only through Business. Treat this as an unresolved code sighting, not a product: configs appear in checkout code for experiments and internal tests that never ship.

ANNOUNCED

Apple cuts employee AI budgets in a cost-control push: some engineers down to $200 a day, with $100 under consideration

Apple has reduced employee budgets for AI usage as part of a broader effort to rein in costs, Bloomberg reported on November 4, citing people familiar with the matter, even as the company pushes staff to use AI more. Last month Apple cut the daily AI spending allowance for some engineers to $200 from $300.

Further reductions are being considered: some engineering teams could go to $100 a day, while non-engineering groups that already operate under a $100 daily cap could fall to as little as $25. Employees who need greater access can seek exemptions from senior executives, according to Bloomberg.

The budgets cover AI 'tokens' — the metered units models charge for — across tools including OpenAI's ChatGPT, Anthropic's Claude and Google's Gemini. Apple did not immediately respond to a request for comment, and the figures are Bloomberg's reporting, not an Apple disclosure; the cuts sit inside a wider cost-control drive under Apple's new chief executive.

ANNOUNCED

Anthropic expands Claude Marketplace to more than 25 partner agents and products, payable from an existing Claude commitment

Anthropic announced on November 4 an expansion of Claude Marketplace, which now lists more than 25 partner agents and products across security, developer infrastructure, creative work and productivity. The catalog is separate from connectors and plugins, which bring outside apps and data into Claude, and from service partners that help companies implement it.

The new listings include Chainguard Harness and Checkmarx One for security, Cyera Agent Guardian and Rubrik Code Guardian for governing Claude deployments and code, and Datadog Agent Observability and Netlify Enterprise for engineering teams. Creative and productivity listings include Runway, Webflow, AirOps, Asana AI Teammates, Glean, monday agents and Zoom Agentic Search.

Eligible enterprise customers can apply part of their committed Anthropic spend to marketplace purchases, covering new purchases and expansions of platforms they already use, so buying does not require a new budget request. Anthropic describes commitment-based buying as a limited preview: organisations use Request to buy or contact their account team to establish eligibility.

ANNOUNCED

Claude Code background subagents now cost about 99% less against your usage limit

Background subagents in Claude Code now use far less of your usage limit, Anthropic's @ClaudeDevs account posted on November 4. A running subagent updates its status line every 30 seconds, and each update used to re-read the subagent's whole conversation; now Haiku 5.5 writes the update from short notes instead, so each update costs about 99% less.

The change targets the status line, not the subagent's work itself — but at 120 updates an hour per running subagent, the old overhead added up for anyone leaving agents running in the background. Anthropic gave no token counts, and the post does not say which Claude Code version carries the change or whether it applies on every provider.

ANNOUNCED

Anthropic pitches Claude as a chip designer to IPO investors — and EDA stocks dip on the claim

Anthropic is telling prospective investors in its upcoming IPO that Claude can help design semiconductors, according to The Information. In recent investor meetings the company said a hardware group formed earlier this year, staffed by about 15 people, is using its own Claude models to speed work on custom silicon — framing AI-assisted chip engineering as an operational edge as it prepares a listing that could value it near $1.5 trillion.

The pitch is that a frontier model can compress design timelines that established chipmakers such as Nvidia and AMD have long treated as fixed. The suggestion alone was enough to move the incumbents of the design-software field: Synopsys shares fell about 2.5% and Cadence about 4.5% after the meetings were reported, with intraday drops of as much as 4% on November 3.

Anthropic has not said it intends to sell Claude's chip-design abilities as a product, and the capability is so far an internal one — the company has reportedly sought proprietary design datasets from electronic-design-automation firms, which is part of why the EDA stocks reacted. Treat the timelines as an investor pitch, not a benchmark: no independent test of Claude-designed silicon has been published.

ANNOUNCED

Anthropic's biology team uses Claude to map the first 3D structure of a bacterial vault particle

Anthropic shared on November 4 how its biology research team worked with Claude to map the first three-dimensional structure of a vault particle from bacterial life. Vaults are hollow, barrel-shaped protein assemblies — among the largest structures in human cells — first identified in 1986, whose biological purpose is still unknown.

The team searched roughly 15 billion predicted protein sequences for distant relatives, finding candidates across 29 major bacterial groups and four archaeal groups, and tested 44 in the laboratory. A phage-derived protein called L05 assembled into a complete vault of 96 copies of itself, about 54 nanometres long and enclosing roughly 25 percent more volume than mammalian vaults.

Claude contributed throughout, from protein searches and DNA design to purification and microscopy analysis: a chromatography approach cut isolation from about three days to roughly 90 minutes, and the first complete cryo-EM model emerged within 34 hours, against weeks conventionally. Anthropic also documents the failures — an additive that interfered with DNA delivery cost two weeks, and Claude produced an inaccurate map after skipping a standard processing step — both corrected under closer supervision.

ANNOUNCED

Anthropic pays a publisher for the first time, licensing AP election data for Claude

Anthropic struck a content and data licensing deal with the Associated Press just before the US midterm elections — the first time it has ever paid a digital news publisher for content, Adweek reported, citing two people familiar with the matter. On Election Night, Claude users asking for polling updates saw up-to-the-minute information provided by and attributed to the AP.

Neither company commented on the record, and the financial terms, duration and scope beyond election data were not disclosed. Anthropic had announced earlier this fall that Claude would show up-to-date results on election night and in the days that follow; the AP deal is what supplied that live data.

The move is a notable strategy shift. Anthropic had argued that its use of publisher content was permitted under fair use and declined the licensing talks OpenAI and Google pursued — the AP was, in fact, the first publisher both of those companies signed. The deal also surfaced days after Anthropic's October 26 S-1 filing, as it prepares for an IPO while facing active copyright lawsuits.

ANNOUNCED

Adobe puts its CX Enterprise Coworker inside ChatGPT, Claude, Gemini and Copilot at once

Adobe announced on November 4 that its CX Enterprise Coworker — an agentic teammate for customer-experience work — is now available inside the AI platforms enterprises already use: ChatGPT as a plugin, Claude, Databricks Genie, Gemini for business and Microsoft Copilot. Availability inside Amazon Quick, IBM and Snowflake is expected to follow.

The Coworker brings agent-led customer-experience workflows into those tools, drawing on a brand's customer data and brand intelligence where the work already happens. Adobe also launched new connectors built as MCP server integrations with partners including Amazon Ads, Databricks Genie, Demandbase, an ads MCP server for Meta campaigns, Microsoft Work IQ, RainFocus and Snowflake Cortex AI, so the agent can take action in those systems rather than only answer questions.

It is another sign of where enterprise agents are heading: the major chat platforms are becoming the front door, and vendors like Adobe compete to be the specialist agent working behind whichever door the customer chose. Adobe did not announce pricing for the Coworker in the release.

2026-11-03

ANNOUNCED

India is now ChatGPT's largest market: OpenAI's India chief says weekly users have doubled since January

India has become ChatGPT's largest market globally by weekly active users, OpenAI India Managing Director Prabhjeet Singh said in an interview with The Economic Times published November 3 — his first media interview since taking the job. Singh said the Indian user base has doubled since January 2026, crediting users weaving the chatbot into everyday work, study and business.

The milestone moves India up from second place. In February 2026, CEO Sam Altman put India at more than 100 million weekly ChatGPT users, behind only the United States. OpenAI has not disclosed a fresh absolute count alongside the new claim, and Singh gave no revenue figures — largest by users is not the same as largest by revenue.

Singh pointed to what comes next: more Hindi and vernacular voice features, Android-first product work, and announcements on UPI-linked payments in the coming weeks. India is also among Codex's largest and fastest-growing markets, he said, and OpenAI has partnered with Tata Consultancy Services on data centre capacity in India through the TCS HyperVault platform, starting at 100 megawatts and scalable to 1 gigawatt.

ANNOUNCED

OpenAI publishes its GPT-5 science paper: case studies across six fields, with the limits documented alongside

OpenAI released 'Early science acceleration experiments with GPT-5' on November 3, a paper co-authored with collaborators at Vanderbilt, UC Berkeley, Columbia, Oxford, Cambridge, Lawrence Livermore National Laboratory and The Jackson Laboratory. It compiles early case studies across math, physics, biology, computer science, astronomy and materials science in which GPT-5 helped researchers synthesise known results in a novel way, run powerful literature reviews, accelerate tough computations and generate novel proofs — and it documents the limitations alongside the wins.

Three cases carry the paper. Immunologist Derya Unutmaz's team spent months on a puzzling change in human immune cells; GPT-5 identified the likely mechanism within minutes from an unpublished chart and suggested the experiment that proved it. Mehtaab Sawhney and Mark Sellke were stuck on the final step of a decades-old Erdős problem, and GPT-5 contributed the idea that completed the proof. Sébastien Bubeck and Christian Coester were testing a common decision-making method used in robotics and routing; GPT-5 produced a clear example showing it can fail and improved a classic optimisation result.

OpenAI's own framing keeps the human in charge: GPT-5 'does not run projects or solve scientific problems autonomously,' but shortens parts of the research workflow in expert hands. The paper is also a positioning document for OpenAI for Science — the company cites a survey in which 60% of US respondents said scientific and medical breakthroughs reach them too slowly — so the case studies are best read as selected examples with named external collaborators, not as a controlled measure of how much faster science now moves.

RELEASED

ChatGPT Voice comes to Finances: talk through your subscriptions and savings out loud

OpenAI formally showed ChatGPT Voice working inside Finances on November 3: after connecting accounts, users can talk through their money out loud — which subscriptions they are paying for, which to keep, and what cancelling the rest could save. OpenAI directs users to chatgpt.com/finances to try it, and its Help Center lists Finances for Free, Go, Plus and Pro users in the US on web, iOS and Android.

Finances itself is not new — it launched as a Pro preview in May, reached Plus and Pro through the autumn, and expanded to Free and Go users on October 2, which is also when voice support first became available to those tiers, according to CryptoBriefing's research. Accounts link through Plaid (12,000+ institutions) with credit data through Experian as a separate connection; the access is read-only, so ChatGPT can see balances and transactions but cannot move money, pay bills or trade.

The November 3 announcement is the showcase, not the first availability — but it moves voice from a quietly shipped capability to the way OpenAI is now pitching the feature. What OpenAI has not published is any date for Finances outside the US.

ANNOUNCED

OpenAI says ChatGPT search runs at one-fifth of Google's volume — and puts weekly users at 1.2 billion

ChatGPT's volume of search-like queries is now roughly one-fifth the volume of Google's, consumer product lead Adam Fry said at an OpenAI-hosted dinner in New York on Monday night, in a fireside chat with Big Technology's Alex Kantrowitz published November 3. Fry's definition is narrow: a search is a user question that makes ChatGPT look something up on the web — current information about people, places or things, plus some finance and health questions. OpenAI did not disclose which Google baseline it divided into, and no independent firm has checked the comparison.

The same conversation put fresh numbers on ChatGPT's scale: 1.2 billion weekly active users, up from 1 billion over the last two months at roughly 100 million a month; 1.5 billion images created a week; and 250 million weekly users of voice input, counting dictation and voice conversations. Gizmodo noted the framing would make ChatGPT the world's second-largest search engine by volume — and that such numbers are exactly what a company says publicly ahead of a planned IPO.

Treat the growth timeline with some care: Fry described the jump from 1 billion to 1.2 billion as happening over the past two months, but The Verge was already reporting the 1.2 billion figure on September 29, so the timing of the increase is unclear. The one-fifth figure is OpenAI's own estimate using its own definition — a company claim about query volume, not a measured share of the search market.

ANNOUNCED

OpenAI says ChatGPT and Codex now do physical engineering: CAD, circuit boards and simulations, with usage up 23-fold

OpenAI said on November 3 that ChatGPT and Codex are being used for mechanical design, electrical engineering and simulation work — creating and revising CAD models, designing printed circuit boards, preparing manufacturing drawings, configuring simulations and reviewing designs. The company announced the work in a thread on X and a same-day post on its site, with Chip Ganassi Racing and Mazda among the partners named.

OpenAI's own teams are the headline user: its hardware and infrastructure groups use the tools for robotics and compute-infrastructure design, and model usage by those teams has risen 23-fold since February, measured in tokens. In one internal example, two employees used GPT-6 Astra to design a test device in under a week as a side project, and a contract manufacturer produced a small prototype batch in under a month.

The bounds matter: the 23-fold figure is usage, not a measured productivity gain, and OpenAI's own prototype still required human correction on tolerances and assembly. This is a move from software production into work where errors carry manufacturing costs, with human review still required — not a claim that the models design hardware unsupervised.