Applied AI insights

Insights

Get Aligned: 16 November 2026

Five under five

Get Aligned, five minutes to stay current, AI for busy teams, 16 November 2026: the AgentAligned computer-head mascot surrounded by app tiles carrying the official ChatGPT (OpenAI), Claude and Grok logos, linked by a gold orbit line.
Conceptual AI illustration featuring the AgentAligned mascot.

AI for busy teams

AI news for busy teams, 9 to 15 November 2026: This week's updates run from your hardest checks to your everyday admin. Claude Fable 5.2 lets an operations or finance lead check a high-stakes document against its sources, Businesses in ChatGPT lets a brand lead claim the company's listing, and Claude Sheets and Whiteboards let a project manager run a tracker and a retro. Sources were rechecked on Sunday 15 November.

Key takeaways

  • Operations, finance and account leads: check one high-stakes document with Claude Fable 5.2, included on Pro and standard Team seats until 16 November at 11:59 pm PT (Tuesday 17 November at 09:59 SAST).
  • Brand or marketing leads: claim your business in ChatGPT's Business Hub and baseline how ChatGPT describes it.
  • Project and account managers: run one project's tracker in Claude Sheets and the next retro on a Whiteboard.

How we chose this week's updates

This edition covers 9 to 15 November 2026. See how we choose Get Aligned.

1. Claude Fable 5.2: which work deserves Anthropic's most capable model?

Announced by Anthropic on 9 November 2026 · Covered by Matt Wolfe and Paul J Lipsky on 9 November, Riley Brown on 11 and 14 November and The AI Advantage on 12 November

Conceptual illustration on a coal background: an open board pack with blank text bars and a small bar chart lies beside a fanned stack of spreadsheet sheets, with fine gold lines linking small gold flags that mark mismatched spots on both, and a gold magnifying glass between them. The official Claude logo sits in the bottom left corner.
Conceptual AI illustration of a board pack checked against its source spreadsheet.

Anthropic says Claude Fable 5.2 is its upgraded model for "the most daunting technical challenges" and is up to 30% faster than Fable 5.1 (Anthropic). Anthropic reports a score of 41.3% at High effort on the Remote Labor Index, which measures paid knowledge work. Early testers quoted by Anthropic describe Fable building a forecasting model from a 10-K with every cell correct and spotting mismatched salary deposits in a loan file. Anthropic still recommends starting most coding and knowledge tasks with Opus 5.5 and moving up to Fable as needed.

Checking a board pack against its source spreadsheet usually falls to one person working line by line late in the day, which is when a wrong figure slips through.

Action

Try with your operations, finance or account lead: In 30 minutes, take one report, proposal or board pack with its source spreadsheet or brief, select Fable 5.2 in the model picker, attach both, then run the prompt below. Use documents you are cleared to share, then hand the agreed fixes to Opus 5.5 to make the edits, as Riley Brown plans with Fable and lets Opus carry out the plan. Useful result: a list you can verify in 15 minutes, with at least one issue you would have missed.

Try this prompt Written for operations, finance and account teams. Paste it into a Claude chat with Fable 5.2 selected and swap the bracketed details.

Check the attached [DOCUMENT] against its source, [SPREADSHEET OR BRIEF], and list every inconsistency: mismatched figures, totals that do not add up, and dates, names or claims the source does not support. For each one, give its location and the evidence from the source. Do not change anything.

Access: Fable is on all paid Claude plans, not Free, included up to 50% of weekly usage limits on Max, Ultra and premium Team and Enterprise seats. Pro and standard Team seats have the same until 16 November 2026 at 11:59 pm PT (Tuesday 17 November at 09:59 SAST), plus one limit reset expiring then for accounts with paid billing before 9 November at 10:00 am PT; after that, Fable runs on usage credits, pay as you go. On Enterprise, it is off until an admin turns it on. Anthropic says "Fable models use your usage limits faster than other Claude models", so watch Settings > Usage.

Explore: Matt Wolfe on availability at 6:30 · Paul J Lipsky on benchmarks at 3:15 · Paul J Lipsky on when to use Fable at 4:15 · Paul J Lipsky on scripts in his voice at 18:16 · Riley Brown on research decks at 15:08 · The AI Advantage on access at 0:47 · The AI Advantage on slide decks at 3:21 · Matt Wolfe on Fable 5.2 at 11:08 · Riley Brown on planning with Fable at 9:30 · Fable launch offers · Fable on your plan

2. Businesses in ChatGPT: how do we claim our business and track mentions?

Announced by OpenAI on 10 November 2026 · Covered by Julian Goldie on 12 November and Matt Wolfe, Paul J Lipsky and The AI Advantage on 13 November

Conceptual illustration on a coal background: a bone-coloured miniature shopfront with a brass key beside it, linked by a gold ribbon to a floating business listing card with a gold location pin and a gold verification tick, with two blank cards behind it. The official OpenAI logo, in white, sits in the bottom left corner.
Conceptual AI illustration of a business claiming its listing in ChatGPT.

OpenAI says businesses can claim their business through Business Hub, in beta, and once verified update key details, add product feeds with pricing and availability, and see aggregated mentions, website clicks and discovery topics (OpenAI). You verify with a business phone number, business email, website domain (a DNS TXT record) or Shopify, and OpenAI reviews the claim and emails you a decision (OpenAI Help Center).

Checking how ChatGPT describes your business usually means someone asking it now and then, while an old description or the wrong website sits there unnoticed.

Action

Try with your brand or marketing lead: In 20 minutes, gather your approved name, description and website and one way to verify, then run the prompt below and save the answer. Sign in at Business Hub, search for your business by name, select your listing and confirm the account that will manage it. Submit only details that match your approved copy. Useful result: a submitted claim and a dated baseline answer to compare after approval.

Try this prompt Written for brand and marketing teams. Paste it into ordinary ChatGPT before claiming and after approval, and swap the bracketed details.

What do you know about [COMPANY NAME] in [CITY OR COUNTRY]? Describe what the business does, list the products or services you associate with it, and give the website address you would cite. Say where you are unsure.

Access: No Ads account or Enterprise agreement is needed. OpenAI says eligible online businesses may see an option to add their business, while local stores and restaurants can currently claim only an existing US listing. Persona identity verification comes before publishing changes or viewing insights, updates are reviewed before they appear, and a blank insight does not mean zero activity. Business Agents, which answer customer questions, are only for eligible US ecommerce and online service businesses; Julian Goldie advises businesses elsewhere to check claim eligibility and not count on them.

Explore: Julian Goldie on verification at 2:15 · Julian Goldie on non-US businesses at 6:00 · Matt Wolfe on Businesses in ChatGPT at 21:36 · Paul J Lipsky on Businesses in ChatGPT at 21:37 · The AI Advantage on the Google Maps comparison at 11:05 · Open Business Hub · OpenAI Help Center on setting up your business · ChatGPT release notes on Business Hub

3. Claude Sheets, Whiteboards and Meetings: can Claude run our tracker and retro?

Announced by Anthropic on 10 and 12 November 2026 · Covered by Paul J Lipsky, Matt Wolfe and The AI Advantage on 13 November and Riley Brown on 14 November

Conceptual illustration on a coal background: a spreadsheet board with gold-highlighted rows and a column of drop-down chips, linked by fine gold lines to a small whiteboard of sticky notes grouped into three clusters. The official Claude logo sits in the bottom left corner.
Conceptual AI illustration of a campaign tracker beside a retro whiteboard.

Anthropic says Claude Sheets (beta) builds spreadsheets with tabs, formulas, formatting and drop-downs from a prompt, and Claude Whiteboards (beta) can group retro notes by theme and draft action items (Claude). Claude Dashboards is now generally available as Claude Data. In Claude Meetings (beta), Claude joins a call as a visible participant, writes notes and action items, and answers questions (Claude). Sheets export to .xlsx or CSV.

Project admin usually means rebuilding the same tracker for every job, then copying retro notes and meeting actions into it by hand.

Action

Try with your project or account manager: In 30 minutes, take one live project's approved scope, owners and dates, run the prompt below, share the tracker and run the next retro on a Whiteboard. Keep it simple: The AI Advantage found Sheets more limited than Google Sheets or Excel, with no pivot tables or macros, but built for Claude to work in. Add Claude to an internal Google Meet call only once everyone agrees to be transcribed. Useful result: every task has an owner and a date.

Try this prompt Written for project and account managers. Paste it into a Claude chat and swap the bracketed details.

For product businesses

Create a campaign tracker in Claude Sheets for our [PRODUCT] launch: tabs for tasks, budget and assets; columns for owner, due date and a status drop-down; budget against actual with a formula for the difference. Use only these details: [APPROVED SCOPE, OWNERS AND DATES]

For service businesses

Create a campaign tracker in Claude Sheets for our [CLIENT NAME] project: tabs for deliverables, approvals and budget; columns for owner, due date and a status drop-down; hours budgeted against hours used with a formula for the difference. Use only these details: [APPROVED SCOPE, OWNERS AND DATES]

Access: Sheets, Whiteboards and Data are on all paid plans, not Free; Sheets is read-only on mobile. All three are on by default on Team; on Enterprise, admins turn on Sheets and Whiteboards, and Data turns on by default on 20 November 2026. Meetings is rolling out on Team (on by default) and Enterprise (admin-enabled), Google Meet only for now, and not for organisations with zero data retention, HIPAA-ready configuration or customer-managed encryption keys. It stays out of calls with outside guests unless an admin allows it.

Explore: The AI Advantage on Sheets at 1:36 · The AI Advantage on Whiteboards at 5:09 · Paul J Lipsky on Claude's new tools at 10:54 · Matt Wolfe on app integrations at 23:05 · Matt Wolfe on Sheets and Meetings at 23:22 · Riley Brown on Sheets and Whiteboards at 0:36 · Riley Brown on Meetings at 5:09 · Get started with Claude Sheets · Get started with Claude Whiteboards · Use Claude Meetings

Also this week

Adobe Firefly adds Quick Actions and brand checks. At Adobe MAX on 10 November, Adobe expanded Firefly as what it calls "the AI creative studio": Quick Actions, Edit Text, Relight and Retouch are generally available now, the Firefly AI Assistant (beta) can extract a brand from a brand guideline PDF and validate images against it, and other tools are in beta or preview (Adobe, Adobe Help). Plans and credits vary; try one Quick Action on an approved asset and brand-check the result.

Gemini 4 Hydrogen reaches free users. Google's fast everyday model reached the Gemini app, including free users via Auto model selection, and Google Search on 12 November; Google says it scores 51 on the Artificial Analysis Intelligence Index (Google).

ChatGPT retires Advanced and Standard Voice on 11 December. OpenAI says GPT-Live becomes the voice experience, though live video and screen sharing stay on Advanced Voice until GPT-Live supports them (ChatGPT release notes).

Abstract illustration on a coal background: a bright white core sends fine lines out to a dozen small floating dark screens, each showing simple charts or waveforms, with small gold nodes along the lines, suggesting many AI agents working in parallel. The official Grok logo, in white, sits in the top right corner.

BONUS

Bonus: one thing to try this week

If there is only one thing you get to do this week, talk through your priorities with Grok Bot by voice, using the new Speed slider (0.5x to 2x) to set a comfortable pace. xAI says Grok Voice 3.0 powers the voice mode; Artificial Analysis ranks it second on its Speech to Speech Index, at 86.9 behind Phonic 1 (High) at 87.1, as checked on 15 November. Some accounts can start a free trial with limited usage. Have a person review anything the Bot drafts before it is sent.

Explore: Introducing Grok Voice 3.0 · Grok Bot changelog · Grok Bot FAQ · Artificial Analysis Speech to Speech Index

What should your team do this week?

Pick the one task that matches work already on your plan, whether a board pack to check, a listing to claim or a tracker to build. Give it an owner, approved inputs and a definition of an acceptable result, and record edits and review time before you decide whether to keep it: that record is the start of building a golden evaluation set.

If the same work keeps stalling, show AgentAligned the bottleneck and we will assess whether a Bottleneck Diagnostic is the right next step.

Frequently asked questions

Is Claude Fable 5.2 included in my Claude plan?

Fable is on all paid Claude plans, not Free. Max, Ultra and premium Team and Enterprise seats get it up to 50% of weekly usage limits; Pro and standard Team seats have it included until 16 November 2026 at 11:59 pm PT (Tuesday 17 November at 09:59 SAST), then on usage credits, pay as you go.

Can a business outside the US use Business Agents?

Not yet. OpenAI limits Business Agents to eligible US ecommerce and online service businesses, but an eligible business located where ChatGPT is available can still claim its business through Business Hub.

Which Claude plans get Sheets, Whiteboards and Meetings?

Sheets and Whiteboards are in beta on all paid plans, not Free; Enterprise admins turn them on. Meetings is rolling out in beta on Team and Enterprise, on Google Meet for now.

Sources

Official company sources

Independent benchmark

Named creators

Source note: product details come from each company's own announcements and documentation, checked on 14 and 15 November 2026. The benchmark ranking reflects the Artificial Analysis table as checked on 15 November 2026. Creator videos are linked for their coverage and demonstrations of each update, not as verification of product facts.

Add Get Aligned to your AI

Paste this into ChatGPT, Claude, Grok or your own assistant so it picks up each new edition.

Every Monday morning, check https://www.agentaligned.agency/insights/feed.xml for the newest Get Aligned edition from AgentAligned. Add it to my daily briefing as three short bullets, one per news item, each with its suggested action and a link to the full article. If you can read my email, look for the Get Aligned newsletter from AgentAligned instead and summarise that.

About the author

Jonathan Oliff co-founded AgentAligned to help in-house brand and marketing teams put AI to work without the hype. He writes the weekly Get Aligned roundup and leads the site's Insights section from Durban. More about how we work.