Amazon vs. AI – Why the Perplexity Lawsuit is a Warning Shot for Agentic Commerce
Nov 12 2025
What’s in store
- Featured story: Amazon vs. AI – Why the Perplexity Lawsuit is a Warning Shot for Agentic Commerce
- Our Take On the News
- Helpful Links & Resources
Amazon vs. AI – Why the Perplexity Lawsuit is a Warning Shot for Agentic Commerce
The world’s largest digital retailer is drawing a line in the sand. Amazon has filed a lawsuit against the AI search engine Perplexity over its Comet Browser AI agents, demanding it cease covert, autonomous purchases on the platform. On the surface, this is a legal battle over access and fraud; strategically, it’s an existential war over the future of digital commerce and Amazon’s ad-revenue model. To directly quote the Perplexity Team’s official response to this lawsuit, “Amazon wants to eliminate user rights so that it can sell more ads right now and partner with AI agents designed to take advantage of users later. It’s not just bullying, it’s bonkers.”
This move by Amazon is the logical, albeit risky, next step in its long-standing ‘content fortress’ pattern. As the market leader, Amazon has the most to lose if a third-party AI agent is allowed to shop autonomously on behalf of a consumer. Amazon is looking to draw lines regarding how consumers can interact with its website, likely because it sees a clear threat regarding its ability to advertise on site. If an AI agent is performing a shopping task on a client’s behalf, navigating, comparing, and purchasing, do display or sponsored product ads actually matter?
Agentic Commerce: Amazon’s Control vs. Walmart’s Capture
Amazon’s defensive stance is a calculated maneuver to preserve its current revenue model and user experience, which is built on the premise of discovery, upselling, and bundling – all of which a hyper-efficient agent threatens. However, this stance creates a major divergence from one of its chief rivals.
The high-level tension is underscored by some irony – Jeff Bezos, Amazon’s founder, is an investor in Perplexity. This tension highlights the core strategic divergence: Amazon is trying to control the shift, while Walmart is attempting to own the new channel. The differing approaches are unsurprising – Walmart is a distant #2 to Amazon in both ecommerce and retail media, and it is willing to take some risks to narrow the gap; Amazon, as top dog, has very little motivation to shake up the status quo.
While Amazon is busy litigating to keep external AI out, Walmart is actively betting on the future of Agentic Commerce. Walmart’s recent partnership with OpenAI integrates a shopping experience, complete with Instant Checkout, directly into ChatGPT. This puts Walmart in a position to capture early market dominance in this emerging paradigm. While this strategy positions Walmart for early market dominance, it will face significant challenges, not least changing entrenched consumer transaction habits.
Three Actionable Strategies for Performance Marketers
This battle presents both a critical challenge and a significant de-risking opportunity for brands and sellers. Your strategy must reflect a two-pronged reality: optimizing for Amazon’s controlled future while simultaneously preparing for the broader, multi-platform agentic world.
- De-Risk: Allocate Budget to Walmart Media – A core function of a performance marketer is to de-risk against single-platform dependency. Allocate incremental ad and commerce budget to Walmart. This hedges against Amazon’s protectionist moves and strategically capitalizes on Walmart’s proactive embrace of third-party agents, which could funnel significant new traffic and sales in the near future.
- Optimize: Win the Internal AI Battle (Rufus) – Amazon isn’t rejecting AI; it’s simply insisting it runs through its own pipes. Its internal AI shopping assistant, Rufus, is projected to generate an incremental $10 billion in revenue. Recognizing this, brands must aggressively adapt product content and strategy to succeed within this internal agentic interface. This means hyper-clean data, optimized keywords, and clear feature descriptions that an AI can easily ingest and recommend.
- Future-Proof: Optimize Data for General LLMs – The agentic future requires a shift from optimizing for search engines to optimizing for general language models. Invest in ensuring your product data is standardized and structured for broad agent consumption. The new competitive edge will be held by brands whose product information is so clean and comprehensive it can be confidently used by autonomous AI agents operating across all major LLMs, independent of any single retailer’s specific ad format.
The Agentic Commerce battle is not just a tech story; it’s a crucial signal for media investment. The core divergence is Amazon protecting its ad-revenue model versus Walmart attempting to gain market share in a potentially new commerce paradigm. As marketers, we must not simply watch the giants clash, but act decisively to structure our data and budgets for the inevitable shift in consumer behavior.
The legal battle between Amazon and Perplexity AI underscores a critical, non-obvious truth for performance marketers: your product data is becoming the new ad currency. When AI agents autonomously search, compare, and purchase on behalf of consumers, the winner won’t be the brand with the highest bid, but the one whose product data is cleanest and most readily consumable by an AI.
This means a fundamental shift in strategy. Investment should move from solely optimizing for platform-specific ad formats toward future-proofing product information for broad agent consumption, independent of any single retailer. Brands with standardized, structured, and comprehensive data will be the first to be surfaced by AI shopping assistants like Amazon’s Rufus and third-party agents integrated with platforms like ChatGPT (via the Walmart/OpenAI partnership). Data agility is your new competitive advantage. | Retail Dive, Perplexity, Modern Retail, Yahoo Finance, Walmart, Fortune
What We’re Tracking
The news stories we’re tracking that are likely to impact advertisers in the month ahead.
U.S. Economy
1. Last week saw oral arguments before the Supreme Court over the Trump Administration’s imposition of blanket tariffs under IEEPA. By way of brief background to what’s brought us here:
- In February the administration declared a new national emergency and imposed fentanyl-related tariffs on Canada, Mexico, and China; in April the administration declared yet another national emergency (this time related to trade deficits) and imposed “reciprocal” tariffs on nearly every country in the world.
- On May 28th the US Court of International Trade found that the President exceeded his statutory authority under IEEPA, vacated the tariff orders, and issued a permanent injunction to halt their collection. On the 29th, the US Court of Appeals for the Federal Circuit stayed the CIT’s injunction, allowing the tariffs to remain in place pending appeal.
- On August 29th, the same Court of Appeals upheld the CIT’s ruling in a 7-4 vote, holding that IEEPA does not authorize the broad tariffs issued by the executive. This ruling was appealed to the Supreme Court.
Which brings us up to the present. Oral arguments were held last Wednesday, and they went …
The prediction markets now reflect a ~80% chance the Court will affirm the lower courts’ rulings and void the tariffs. In such a scenario the Administration would have other tools at its disposal to tax international trade, but it would become more difficult to tax goods having nothing to do with national security. | Kalshi
2. Tariff repeal – i.e. a tax reduction – might be a welcome reprieve, because consumers are gloomier than they’ve been in a long time. The November sentiment index dropped 3.3 points to 50.3, just above a June 2022 reading of 50 that was the weakest in University of Michigan data back to 1978.
The drop in sentiment was broad-based across age, income and political groups; among Democrats and political independents, confidence slid to the lowest in data back to 1984. In contrast to consumers’ anxieties in recent years, this does not appear to be about inflation, as inflation expectations have actually moderated; it’s now the labor market that is causing angst, as consumers “… anticipate that labor markets will continue to weaken in the future and expect to be personally affected.” | Bloomberg
3. Speaking of the labor market, we’re in something of a data blackout owing to the federal government shutdown. But the other indicators we can reference are not encouraging – ADP data indicate net job creation of just 42k in October.
Looking at broader trends, namely the 12-month change in private-sector employment, it seems clear the labor market is losing strength, with job creation far down from a couple of years ago:
The state of the labor market is, by all accounts, weighing on the collective minds of the Fed more and more. Having already implemented two quarter-point rate cuts this year, markets are pricing in a ~75% chance of further interest rate cuts in December:
Now that the federal shutdown has ended, we’ll soon get a better picture of where the labor market stands. | Bloomberg, NYT
Tech Giants & Platforms
- Let’s do a roundup of the major earnings announcements, shall we?
- Meta announced robust 26% YoY revenue growth, fueled by rises in impression availability and in price-per-ad, two metrics that almost always move in opposite directions. But its stock slid after it told investors that it plans extremely aggressive capital expenditures on AI infrastructure. | WSJ, Stratechery
- Google announced YoY revenue growth of 16%, with its advertising business up almost 13% (the other major component is cloud, where growth was impressive). Within advertising, search and YouTube were both up 15%, while the network business (buying and selling on the open internet) shrank by almost 3%. | WSJ, Stratechery
- Amazon announced headline YoY revenue growth of 13%, while Amazon Ads grew at a 24% clip to reach almost $18 billion for the quarter. The company flagged Thursday Night Football viewership averaging 15.3 million, a 16% increase over last season; it also cited “strong growth in demand-side platform (DSP) offerings.” | WSJ
- Apple announced 8% YoY growth, with growth in its services business of 15%. This is the part of Apple that includes advertising, the App Store, iCloud, Apple Pay, its subscription services, and several others. The company achieved a notable milestone in surpassing $100 billion in services revenue for the year; it’s estimated that advertising makes up about 10% of the services business. | WSJ
- Reddit posted blistering 68% revenue growth, however its audience growth is decelerating – US user growth was 7% this past quarter, down from 12% the quarter prior. | eMarketer
- Spotify posted 12% revenue growth (on a constant currency basis), but its ad business did not grow at all. The bright spot is user engagement, where the company saw monthly active users climb 11% to 713 million. | WSJ
Media & Advertising
1. Another season of Major League Baseball is officially in the books, and the league can feel pretty good about viewers’ interest in America’s pastime (sorry Blue Jays fans, but it’s ours!). The postseason averaged 6.33 million viewers per game, up 28% over last year, making it the most-watched playoffs since 2017. Viewing figures were helped by a record seven winner-take-all games across the wild card, divisional series, league championship series, and the World Series. These strong postseason numbers followed comparable growth in regular season viewing, which rose ~20% this year.
The World Series averaged 16.1 million viewers, +2% over last year’s Dodgers – Yankees series; this was a pleasant surprise, as Dodgers – Yankees – a matchup between glamour franchises representing the two largest media markets in the country – was seen as a best-case-scenario for viewership. And indeed, viewership was down 18% YoY through five games, but the intense drama of games 6 and 7 drove the average above 2024. The classic that was game 7 drew over 27 million viewers across Fox platforms, making it the largest MLB audience since game 7 in 2017.
Game 7 ranks as easily the most-watched non-NFL sporting event on any network — surpassing the college football and basketball national championship games and Game 7 of the NBA Finals; it was the fourth time in six years that the World Series pulled more viewers than the NBA’s showpiece. | SMW, SMW, SMW
2. As we’ve been saying for some time, FAST (free ad-supported streaming TV) is officially mainstream: US FAST viewing hit 1.8B hours in August (+43% YoY), and ad tiers are now core to streamers’ strategy—45% of Netflix viewing is ad-supported, a 30%+ rise YoY, with similar gains across Prime Video, HBO Max, Discovery+, and Disney+. With FAST tiles built into smart-TV home screens (Samsung TV Plus, Google TV), it increasingly feels like “just TV”, shifting the battle from acquisition to retention and curation. For advertisers, scale is abundant but monetization lags: fill rates are slipping as supply outpaces demand. The move now is a hybrid CTV mix—use FAST for efficient reach and first-party engagement signals while keeping paid CTV for exclusives and high-value audiences. We have seen robust growth in terms of audience and supply with our FAST partners, and their CPMs, for the aforementioned reasons, remain extremely attractive. | eMarketer
Helpful Links & Resources
- Paramount reported flat revenue, with 17% growth in streaming and a 12% decline in linear
- High School Football can have stream ads too
- Hulu + Live TV and Fubo Merger
- YouTubeTV flexes muscles in showdown with Disney
- Trade Desk partners with Shopsense for shoppable inventory across open web
- Netflix’s Ad Tier Has Almost Half of Its Household Viewing Hours
- While Paramount eyes the whole thing, Netflix is sizing up WBD’s studio and streaming units
- TikTok is launching a podcast network with iHeart Media
You Might Be Interested In
