# Amazon, Muse, and the Layer an Agent Can Skip — InvestMoat Research

_The tape read Amazon blocking Meta's Muse as the store losing its interface, and CoreWeave and Nebius as AWS losing the workload. Both stories sort on a layer an agent or a GPU renter can skip. The fortress is the layer they cannot._

_Published: September 22, 2026 · Last reviewed: September 22, 2026 · Canonical page: https://investmoat.com/research/amazon-muse-and-the-layer-an-agent-can-skip_
_Tags: Big Tech, Platforms, Business Models, AI Monetization_
_Covers: AMZN, META, SHOP, WMT, MSFT, GOOGL, CRWV, NBIS_

## Summary

Amazon blocked Meta's Muse from amazon.com on September 20–21, 2026, after asking Meta to take the store out of the agent. The market heard a fortress cracking. The framework already marks Amazon's learned-interface pillar as the weak one — Alexa is table stakes — and rates the transaction, the marketplace network, the purchase-data exhaust, and the AWS record as the durable half. Muse skips the storefront. It does not skip Fulfillment by Amazon, the Prime badge, or the Buy Box. The same skip test applies to the other dying-Amazon story: CoreWeave and Nebius are overflowing GPU racks whose biggest counterparties are the hyperscalers themselves, not a second IAM. AWS printed $42.2B of sales in Q2 2026, growing 37%, against CoreWeave's $2.575B and Nebius's $582.3M in the same quarter. An agent that reorders paper towels still has to clear a warehouse. A training run that bursts onto rented GPUs still has to come home to a cloud account.

## What would prove this wrong

Advertising services growing below 10% year-over-year for two consecutive quarters while third-party seller services stall below 8% would show discovery and the marketplace leaving with the agent. An advertising slowdown traced to a remedy in the FTC and 22-state ad-auction suit filed August 31, 2026 does not trip it — that is a pricing event, not a discovery one. A second trip: AWS year-over-year growth falling below 25% for two consecutive quarters while remaining performance obligations fail to grow from the $496B printed in the Q2 2026 10-Q — overflow becoming the demand, not spare capacity on a still-compounding record.

_Status as of September 22, 2026: **holding**._

---

**The two Amazon-is-dying stories in one grid: the agent that wants the storefront, the open checkout and the store grid, the other hyperscaler records, and the GPU landlords. Scores are live — moat, growth, valuation and composite recompute from each stock's JSON against the current price.**

| Ticker | Name | Moat | Growth | Valuation | Composite | Note |
| --- | --- | --- | --- | --- | --- | --- |
| **Two-record target** |  |  |  |  |  |  |
| [AMZN](https://investmoat.com/stocks/amazon) | Amazon.com Inc. | 93 | 84 | 80 | 89 | AWS $42.2B +37%; ads $19.8B +26%; 3P $46.8B; blocked Muse Sept 20–21 |
| **The agent that wants the interface** |  |  |  |  |  |  |
| [META](https://investmoat.com/stocks/meta) | Meta Platforms Inc. | 81 | 74 | 79 | 80 | Muse launched Sept 8; Stripe Link live; Shop Pay named; no Muse revenue line |
| **Open checkout and the store grid** |  |  |  |  |  |  |
| [SHOP](https://investmoat.com/stocks/shop) | Shopify Inc. | 80 | 91 | 71 | 83 | The rail Muse named; Shop Pay coming, per Meta's September 8 post |
| [WMT](https://investmoat.com/stocks/walmart) | Walmart Inc. | 74 | 68 | 71 | 71 | The store grid an agent can route around the app and still has to clear |
| **Other hyperscaler records** |  |  |  |  |  |  |
| [MSFT](https://investmoat.com/stocks/msft) | Microsoft Corp. | 90 | 79 | 77 | 84 | The enterprise record enterprises already dual-run with AWS |
| [GOOGL](https://investmoat.com/stocks/google) | Alphabet Inc. | 85 | 78 | 75 | 81 | A cloud record on a search company — the other hyperscaler, not a neocloud |
| **Overflow landlords** |  |  |  |  |  |  |
| [CRWV](https://investmoat.com/stocks/crwv) | CoreWeave, Inc. | 35 | 85 | 78 | 62 | Q2 revenue $2.575B +112%; backlog ~$104B; net loss $626M |
| [NBIS](https://investmoat.com/stocks/nbis) | Nebius Group N.V. | 35 | 84 | 50 | 51 | Q2 revenue $582.3M +454%; not a second IAM |

_Valuation and composite above use each asset's static valuation score; the live site recomputes them against the current market price._

Two Amazon-is-dying stories landed in the same week, and they sound like one argument. Meta launched Muse on September 8. Amazon blocked it from amazon.com on September 20–21 after asking Meta to take the store out of the agent. Users who sent Muse shopping hit an anti-bot wall. The tape heard a fortress cracking: if Amazon has to lock the door, the storefront is already lost. The other story has been running all year. Companies are supposedly leaving AWS for CoreWeave and Nebius, because a new cloud that rents NVIDIA racks can offer the same service the old one does.

**Both stories sort on a layer an agent or a GPU renter can skip.** The [MAG7 piece](/research/amazon-and-the-moat-the-market-still-calls-retail) already argued that Amazon is the system of record in physical commerce and in a hyperscale cloud at the same time. The [Muse note](/research/meta-muse-and-the-whatsapp-test) already argued that a personal agent sorts on habit plus checkout, and that Alexa is this category without the habit. This page is the skip test those two pieces imply and did not run: which layer did Muse actually threaten, and which layer do the neoclouds actually replace.

## Amazon blocked the skip. It did not block the purchase

**Three weeks around the block — what was sued, what shipped, what was refused, and which Amazon agent was already compounding.**

| Date | What happened | What it tests |
| --- | --- | --- |
| August 31, 2026 | FTC and 22 states sue Amazon, alleging undisclosed "soft reserve" surcharges in its sponsored-ad auctions since 2019. Amazon denies it | Whether the ads line can slow for a reason that is pricing, not the agent — the confound the falsifier has to exclude |
| September 8, 2026 | Meta launches Muse in the US: own app, muse.ai, WhatsApp; checkout via Link by Stripe; Shop Pay named as coming | Whether a personal agent can close a purchase Meta observes |
| September 20–21, 2026 | Amazon blocks Muse after asking Meta to exclude the store. Popup: unauthorized AI agent violates the Conditions of Use | Whether the storefront can refuse the skip |
| Q2 2026, printed July 30 | Amazon folds Rufus and Alexa+ into Alexa for Shopping; active users close to doubling, interactions up over 5× YoY | Whether Amazon wants an agent on the interface — it does; it wants its own |

_Figures as of September 21, 2026. Source: [FTC, States Sue Amazon Over Secret Ad Surcharge Scheme](https://www.ftc.gov/news-events/news/press-releases/2026/08/ftc-states-sue-amazon-over-secret-ad-surcharge-scheme); [Introducing Muse: The World's First Personal AI Agent Built for Everyone](https://about.fb.com/news/2026/09/introducing-muse-personal-ai-agent/); [Amazon shows Meta's Muse AI shopping agent the door](https://www.theregister.com/ai-and-ml/2026/09/21/amazon-shows-metas-muse-ai-shopping-agent-the-door/5297777); [Amazon.com Announces Second Quarter 2026 Results](https://ir.aboutamazon.com/news-release/news-release-details/2026/Amazon-com-Announces-Second-Quarter-Results/default.aspx)._

Amazon's own statement, as The Register printed it, is an opt-in rule borrowed from food delivery and online travel: third-party applications that buy from other businesses should operate openly and respect a provider that says no. The same write-up is honest about the cash register behind the rule. Amazon generated more than $68B of advertising revenue last year because shoppers still walk the store. An outside agent that compares, decides, and pays from a Stripe one-time card never sees the sponsored placement. That is a real threat to the interface and to the ads line that sits on it. It is not, by itself, a threat to the warehouse.

**Amazon Q2 2026 — the store lines Muse would have to break, and the cloud line the neocloud story would have to break.**

| Line | Q2 2026 | What it tests |
| --- | --- | --- |
| Advertising services | $19.8B, +26% YoY | Whether discovery still happens on Amazon's surface |
| Third-party seller services | $46.8B, +16% YoY | Whether the marketplace network is still the place a seller has to be |
| North America operating income | $9.1B, vs $7.5B | Whether density is still earning its cost |
| AWS sales | $42.2B, +37% YoY — fastest in 18 quarters | Whether the cloud record is still compounding |
| AWS remaining performance obligations | $496B; 6.4-year weighted-average life | Whether the cloud record is contracted, not hoped |

_Figures as of July 30, 2026. Source: [Amazon.com Announces Second Quarter 2026 Results](https://ir.aboutamazon.com/news-release/news-release-details/2026/Amazon-com-Announces-Second-Quarter-Results/default.aspx); [Amazon.com Form 10-Q for the quarter ended June 30, 2026 — remaining performance obligations](https://www.sec.gov/Archives/edgar/data/1018724/000101872426000026/amzn-20260630.htm)._

The ads line is the one the block is written to protect. It is also the one that was still accelerating in the last print — +26% on $19.8B, in the same quarter third-party seller services at $46.8B outsold AWS. A ranking that starts from the popup will call the storefront dying. A ranking that starts from those two lines will say the interface is the thing Amazon is willing to police, because the network and the transaction are still printing. The [habit-plus-checkout note](/research/meta-muse-and-the-whatsapp-test) already put Amazon's learned-interface pillar on the weak side of the Muse cohort. The block is that cell defending itself. It is not a downgrade of the other four.

> **The skip test** — Ask of any agent, and of any new cloud: which layer does this product skip, and which layer does it still have to call? Skipping the storefront is not skipping Fulfillment by Amazon. Skipping a training cluster is not skipping IAM. A moat dies when the layer you cannot skip becomes skippable. It does not die when the company locks a door on a layer the framework already marked as weak.

## Eight names, one skippable layer

**The five pillars the two dying-Amazon stories actually use. Statuses are read live from each stock's analysis.**

| Ticker | Learned Interfaces | Transaction Embedding | System of Record | Network Effects | Proprietary Data |
| --- | --- | --- | --- | --- | --- |
| **Two-record target** |  |  |  |  |  |
| [AMZN](https://investmoat.com/stocks/amazon) | weakened | strong | strong | strong | strong |
| **The agent that wants the interface** |  |  |  |  |  |
| [META](https://investmoat.com/stocks/meta) | strong | strong | intact | strong | strong |
| **Open checkout and the store grid** |  |  |  |  |  |
| [SHOP](https://investmoat.com/stocks/shop) | intact | strong | intact | strong | strong |
| [WMT](https://investmoat.com/stocks/walmart) | weakened | intact | na | strong | intact |
| **Other hyperscaler records** |  |  |  |  |  |
| [MSFT](https://investmoat.com/stocks/msft) | intact | strong | strong | strong | strong |
| [GOOGL](https://investmoat.com/stocks/google) | strong | strong | intact | strong | strong |
| **Overflow landlords** |  |  |  |  |  |
| [CRWV](https://investmoat.com/stocks/crwv) | na | weakened | na | na | na |
| [NBIS](https://investmoat.com/stocks/nbis) | na | weakened | na | na | na |

The grid does the work the popup cannot. Amazon is strong on the transaction, the record, the network and the data, and the learned-interface pillar is the one the framework marks as weak. That is the cell Muse attacks. Walmart's learned-interface pillar is marked the same way — an agent can shop Kroger, Target, or Amazon through one chat window — and Walmart's fortress is the store grid, not the app. Shopify is the name that invited the skip: Shop Pay is the rail Meta named on September 8, and the transaction-embedding pillar is the one Shopify wants an agent to call. CoreWeave and Nebius reach a durable mark on none of the five. That is the framework saying, without sentiment, that a rented GPU against a loan is not a cloud record, however large the backlog attached to it. The [hyperscaler-build piece](/research/the-odds-on-alphabets-ai-build) already grouped those two as borrowing to build for others. This page is the competitive claim that grouping implies: overflow is not displacement.

## A $42B record does not lose to a $3B landlord

**Q2 2026 cloud sales, same quarter. The neoclouds are real businesses. They are not a second AWS.**

| Name | Q2 2026 | What it is |
| --- | --- | --- |
| AWS | $42.2B, +37% YoY; $496B remaining performance obligations | The cloud system of record — S3, IAM, the account, GovCloud |
| CoreWeave | $2.575B, +112% YoY; ~$104B revenue backlog; GAAP operating loss $49M; net interest $640M | Overflow GPU rental, financed |
| Nebius | $582.3M, +454% YoY; adjusted EBITDA $236.2M; GAAP net loss from continuing operations $190.4M | Smaller overflow, same job |

_Figures as of August 12, 2026 — latest print in the table (Nebius Q2). Source: [Amazon.com Announces Second Quarter 2026 Results](https://ir.aboutamazon.com/news-release/news-release-details/2026/Amazon-com-Announces-Second-Quarter-Results/default.aspx); [CoreWeave Reports Strong Second Quarter 2026 Results](https://www.sec.gov/Archives/edgar/data/1769628/000176962826000362/coreweave2q26earningspress.htm); [Nebius reports second quarter 2026 financial results](https://assets.nebius.com/assets/72a8c258-bbb7-4df7-ab9d-8698f6cb88fc/PR.pdf)._

CoreWeave in one quarter is about a sixteenth of AWS. Nebius is about a seventy-second. Those ratios are not a smear. They are the scale at which 'a lot of companies are leaving AWS' would have to print if it were happening. What the neocloud prints actually show is the opposite customer: Microsoft, Meta, and the frontier labs buying spare training capacity from a specialist while they keep the identity plane, the object store, and the compliance perimeter where they already are. Microsoft declining a CoreWeave expansion and signing with another neocloud — the fact the CoreWeave analysis already carries — is the tell. Overflow is interchangeable. IAM is not. The [Palantir piece](/research/palantir-and-the-layer-an-agent-has-to-call) ran this test on software: the agent has to call the ontology. The cloud version is the same sentence with a different noun. The training run can call CoreWeave. The account cannot.

Jassy told investors AWS is capacity-constrained through 2027. A capacity-constrained record leaking overflow to a specialist is how you would expect this cycle to look. It is not how a degrading record looks. A degrading record loses the contracted book, or it loses growth while the specialist takes the enterprise account. AWS grew 37% — the fastest in 18 quarters — and the 10-Q carried $496B of remaining performance obligations with a 6.4-year weighted-average life. That book doubled in two quarters. The [build piece](/research/the-odds-on-alphabets-ai-build) already warned that a material share of hyperscaler backlog is circular: Amazon funds Anthropic, Anthropic commits AWS spend. Circularity is a growth and ROIC fact. It is not evidence that Nebius replaced S3.

## The agent still settles. The rack still rents

- **Learned interface — the skippable layer.** Alexa, the app, the search box, the sponsored module. Muse's browser is built to walk around that surface. The framework already rates the cell as weak on Amazon and on Walmart. Amazon's own Alexa for Shopping is the in-house version of the same skip. The block is Amazon refusing to let someone else's agent sit in a cell it already knows is not the moat.
- **Transaction embedding — the layer Muse still has to call if the item ships from Amazon.** 1-Click, Prime, FBA, the Buy Box. Automating the shopper does not remove the settlement. A Stripe one-time card can pay. It cannot put the SKU in a same-day node the seller does not run.
- **System of record — two of them, and neither is a chat window.** The store record is the order, the catalog, the last node that touched the SKU. The cloud record is S3, IAM, the account, the $496B book. An agent that reorders paper towels still clears the first. A training job that bursts onto CoreWeave still comes home to the second.
- **Network effects — the marketplace a seller does not dual-run.** Third-party seller services at $46.8B in a quarter is the network as revenue. A household does not hold two Prime badges. A seller does not split FBA across two fulfillment graphs that both confer the badge. GPU rental has no such loop: CoreWeave's and Nebius's analysis both mark the network cell as not applicable, because compute is fungible across landlords.
- **Proprietary data — the exhaust the ads line bills.** Purchase intent and last-mile telemetry are what advertising at $19.8B, +26%, is charging for. Muse that never walks the store never feeds that exhaust. That is the cash reason for the block. It is also why the block, if it holds, protects the data pillar rather than proving it already died.

The cloud half of the same list is shorter because the neoclouds do not claim the other cells. They claim price and speed on a GPU generation. That is a real product. It is the product the [build piece](/research/the-odds-on-alphabets-ai-build) already sorted as borrowed capacity with no cash engine to retreat to. Microsoft and Alphabet sit in this cohort as the control: they are the other places an enterprise already dual-runs a record. Dual-running Azure and AWS is common. Dual-running AWS and CoreWeave as the identity plane is not a sentence anyone in coverage has a print for.

## Shopify wants the skip, and Amazon already runs it the other way

The other side of the store half is not a straw man, and it is not CoreWeave. It is Shopify, and a reader who thinks Muse is how Amazon's ads line dies is reading a real architecture, not a headline. Meta's own September 8 post names Shop Pay as coming, next to Link by Stripe. That is an invitation. Shopify's transaction-embedding and network pillars are the ones the framework rates as strong because the merchant already made Shopify the checkout, the wallet, and the app store — and the company has spent a year telling anyone who will listen that 'Commerce for Agents' is the point of the platform. An agent that can see the catalog, compare, and pay on Shop Pay never needs amazon.com. If that path takes the reorders that today hit 1-Click, the ads line is not a protected exhaust. It is a residual of a storefront people stopped opening.

Amazon has already conceded the mechanism in the other direction. Buy for Me, as The Register notes, sends Amazon's own agent onto other retailers' sites, identifies itself, and lets those retailers opt out. That is the same opt-in rule Amazon is now enforcing against Muse, pointed outward. A reader who calls that hypocrisy is not confused. They are noticing that Amazon wants to be the agent that skips other people's storefronts and the storefront that other people's agents cannot skip. If the category standard becomes the Shopify invitation rather than the Amazon wall — if a federal appeals court that already lifted a Perplexity bar in August keeps lifting them — then the wall is a delay, not a moat, and the learned-interface cell is not the only one that moves. Discovery leaving is how you would see it: advertising growth rolling over while third-party seller services hold, because the seller still needs FBA and the buyer no longer needs the search box.

Walmart is the physical version of the same warning. The framework already writes the app down and the store grid up. An agent that shops around can still have to clear 10,900 stores for groceries that have to arrive in hours. That is why Walmart belongs on this page as a control, not as a victim. It is also why the Amazon store half is not automatically safe: the items that do not need a same-day node — the SKU a Shop Pay merchant can ship from their own warehouse — are the ones an agent can take with it. The MAG7 piece treated the warehouse as the increment that Microsoft does not have. The increment only applies to the SKUs that use it.

The cloud half of the other side is Microsoft, and it is the same honest counter-case the MAG7 piece already ran. Active Directory is a worse thing to leave than an S3 bucket. Enterprises already multi-cloud the workload. If the next wave of training and inference is what Azure and the neoclouds split between them, AWS can keep the record and still lose the growth that is paying for the $220B capex year. CoreWeave's own chief executive called Q2 an inflection toward enterprise, not only frontier labs. Take that sentence at full strength. A specialist that starts as overflow and ends as the default training cloud for the Fortune 500 is exactly the path AWS itself ran from 2006. The print that would make that true is not a $2.6B quarter. It is an enterprise identity story CoreWeave and Nebius do not yet have, and that Microsoft already does. Anyone who wants to invert this page has to argue that GPU overflow becomes the record. That argument is coherent. It is also the argument the neoclouds have to prove, not the one the Q2 scale table currently supports.

## Do not underwrite the popup. Underwrite the line the skip would break

This page is not a claim that Amazon is cheap, and it is not a claim that Muse will fail. The [Muse note](/research/meta-muse-and-the-whatsapp-test) already said the agent is two weeks into a habit test and has no named revenue line; nothing in a weekend block changes that. The [MAG7 piece](/research/amazon-and-the-moat-the-market-still-calls-retail) already said durability is the argument and the price is a different page. The job here is narrower. The market took a locked door and a pair of GPU landlords and sorted Amazon as a company whose moat is slowly becoming a service anyone can offer. The five-pillar grid sorts the same facts as a weak interface being policed and a cloud record still compounding, with overflow sitting in the cells that have no record to lose.

> **What would break this** — Three things, in order. **One:** advertising services below 10% year-over-year for two consecutive quarters while third-party seller services stall below 8% — discovery and the marketplace leaving with the agent, which is the Shopify invitation winning. A slowdown that traces to a remedy in the FTC's August 31 ad-auction suit does not count; that is Amazon losing a pricing lever, not the agent taking the shopper. **Two:** AWS growth back below 25% for two consecutive quarters while remaining performance obligations fail to grow from $496B — the contracted book turning into a cliff, which is the only version of the neocloud story that is a moat event. **Three:** disclosed Muse or Shop Pay volume becoming a named commerce path that Amazon's own Alexa for Shopping does not match, which would mean the interface cell is not only weak but no longer Amazon's to police.

The lesson generalises past this weekend. **When a company locks a door, check whether the door is on the fortress or on the room it already knew was a hallway.** Amazon locked the storefront. Muse still has to call the warehouse if the item is an Amazon item. CoreWeave still has to call a hyperscaler account when the job is more than a rack. The MAG7 factor hears a threat in any new product that says it can do the same job. The skip test asks whether the job it can do is the one that was ever the moat.

## Sources

1. [FTC, States Sue Amazon Over Secret Ad Surcharge Scheme](https://www.ftc.gov/news-events/news/press-releases/2026/08/ftc-states-sue-amazon-over-secret-ad-surcharge-scheme) — Federal Trade Commission, August 31, 2026 (regulator)
2. [Introducing Muse: The World's First Personal AI Agent Built for Everyone](https://about.fb.com/news/2026/09/introducing-muse-personal-ai-agent/) — Meta, September 8, 2026 (company-site)
3. [Amazon shows Meta's Muse AI shopping agent the door](https://www.theregister.com/ai-and-ml/2026/09/21/amazon-shows-metas-muse-ai-shopping-agent-the-door/5297777) — The Register, September 21, 2026 (third-party)
4. [Amazon.com Announces Second Quarter 2026 Results](https://ir.aboutamazon.com/news-release/news-release-details/2026/Amazon-com-Announces-Second-Quarter-Results/default.aspx) — Amazon, July 30, 2026 (press-release)
5. [Amazon.com Form 10-Q for the quarter ended June 30, 2026 — remaining performance obligations](https://www.sec.gov/Archives/edgar/data/1018724/000101872426000026/amzn-20260630.htm) — Amazon / SEC, July 31, 2026 (filing)
6. [CoreWeave Reports Strong Second Quarter 2026 Results](https://www.sec.gov/Archives/edgar/data/1769628/000176962826000362/coreweave2q26earningspress.htm) — CoreWeave / SEC, August 11, 2026 (filing)
7. [Nebius reports second quarter 2026 financial results](https://assets.nebius.com/assets/72a8c258-bbb7-4df7-ab9d-8698f6cb88fc/PR.pdf) — Nebius Group, August 12, 2026 (press-release)

---

Scores are computed deterministically from each asset's data file: composite = Moat^0.40 × Growth^0.30 × Valuation^0.30, scaled to 0–100.

Full analyses for every name above: https://investmoat.com/stocks. Site index for agents: https://investmoat.com/llms.txt

InvestMoat is an open-source research and education framework. Nothing here is financial advice.