Goldman's AI Buy List Leans on Adopters, Not Chipmakers

Goldman Sachs published an AI stocks to buy list arguing the biggest payoff may soon move beyond chipmakers, naming CSGP, DLTR, EBAY, AJG, BRO, AXON, TTD, ABNB, BA, EXPE, TGT and SCHW; all 11 of those with live quotes were lower on Aug. 17, 2026.

Goldman Sachs has put out an artificial-intelligence buy list that does something most AI screens still refuse to do: it looks past the companies selling the hardware. The bank's framing, as reported by GuruFocus, is that the biggest payoff from AI may soon move beyond the chipmakers and toward the businesses that deploy the technology inside their own operations.

The names attached to the list are a deliberately unglamorous set: CoStar Group (CSGP), Dollar Tree (DLTR), eBay (EBAY), Arthur J. Gallagher (AJG), Brown & Brown (BRO), Axon Enterprise (AXON), The Trade Desk (TTD), Airbnb (ABNB), Boeing (BA), Expedia (EXPE), Target (TGT) and Charles Schwab (SCHW). Not a semiconductor among them.

Why the argument shifts from sellers to buyers of AI

The logic behind an "adopter" list is straightforward, even if the timing is contentious. Chip and infrastructure vendors captured the first wave of AI economics because they were paid up front, in cash, by customers competing for scarce capacity. That trade has been the market's dominant story for years and is now heavily owned. The second wave, in this telling, accrues to the companies that turn the spending into lower unit costs, better pricing, faster underwriting or higher conversion rates — margin gains that show up in operating income rather than in a backlog.

That is a harder thesis to prove and a slower one to trade. A chip order is visible; a claims-handling workflow that needs 15% fewer人 hours is not. It is also why the list skews toward businesses with large, repetitive, document-heavy or matching-heavy processes.

How the twelve names group together

Read as a set, the list clusters into four recognizable buckets.

• Data and information businesses: CoStar Group sits on a proprietary commercial-property dataset, the kind of asset that becomes more valuable when models can query and summarize it at scale.
• Insurance distribution: Arthur J. Gallagher and Brown & Brown are brokers, not underwriters. Their costs are people, paperwork and placement — the clearest near-term target for automation in financial services.
• Consumer marketplaces and travel: eBay, Airbnb and Expedia all live or die on search, matching and personalization. Better ranking translates directly into take rate.
• Retail, industrial and brokerage: Dollar Tree and Target are inventory, labor-scheduling and supply-chain problems at enormous scale. Boeing is a manufacturing and engineering-documentation problem. Charles Schwab is a service-desk and advice-delivery problem. Axon Enterprise, with its evidence and report-writing software, arguably straddles vendor and adopter.

What links them is not AI revenue. It is AI as a cost line that eventually bends the other way.

Monday's tape did not cooperate

The list landed on a soft session. As of the last trade at 17:35 GMT on Aug. 17, 2026, the S&P 500 tracker SPY was at $773.47, down 0.37%, with the Nasdaq 100 proxy QQQ at $730.01, off 0.14%, and Dow tracker DIA at $533.88, down 0.54%. Every one of the 11 named stocks carrying a live quote was lower on the day.

The Trade Desk was the weakest, at 13.39, down 5.30% from a prior close of 14.14 and near the bottom of a 13.33–13.98 range — a reminder that the programmatic advertising name has already been repriced hard, whatever the AI overlay. CoStar Group fell 3.27% to 31.32. Expedia dropped 3.06% to 322.51, and Airbnb slid 2.03% to 180.32, the travel pair moving together as usual.

Boeing lost 2.06% to 226.90. Among the brokers, Brown & Brown fell 2.24% to 68.96 and Arthur J. Gallagher 1.35% to 247.81. Target eased 1.50% to 152.16, Axon 1.07% to 606.29, eBay 0.75% to 102.37, and Dollar Tree was effectively flat, down 0.05% at 129.33. Goldman Sachs itself (GS) was the outlier in the group, up 0.66% at 1046.24.

Note the prices as quoted: the licensed feed supplying these figures does not specify the exchange or the currency for the individual tickers, so they are given here exactly as delivered.

What this list is actually betting on

Three things have to be true for an adopter basket to beat the infrastructure trade.

First, the productivity has to be measurable. Investors will not pay for a narrative about efficiency; they will pay for a gross-margin line or a headcount-to-revenue ratio that visibly improves over consecutive quarters. Insurance brokers and brokerages report enough operating detail for that to be testable.

Second, the savings have to stay with the company rather than being competed away. In marketplaces and travel, where rivals adopt the same tools from the same vendors, better matching can end up passed straight to the consumer as lower prices. That is good for users and neutral for shareholders.

Third, the spending has to stop rising faster than the benefit. Adopters are also buyers — of cloud, of licenses, of data. A company that pays vendor-level prices for tools that deliver commodity gains has simply added a cost.

What to watch next

The practical tests arrive with earnings, not with lists. Watch for retailers quantifying shrink, markdown or labor-scheduling gains; brokers disclosing revenue per employee; marketplaces breaking out conversion or take rate; and any of these companies raising capital-expenditure guidance for internal AI without pairing it to a target return.

Also worth watching is the composition risk in a screen like this. A basket that spans discount retail, commercial real-estate data, insurance broking, aerospace and online travel is exposed to consumer spending, property markets, rate expectations and the aircraft cycle all at once. On Monday those factors pulled in the same unhelpful direction, which is how a group with no shared industry can still fall together. AI may be the reason the names were selected; it will not be the only reason they move.

Frequently asked questions

What is Goldman Sachs actually recommending?

Goldman Sachs published an AI stocks to buy list built around the idea that the biggest payoff from artificial intelligence may soon move beyond chipmakers. The named stocks are CoStar Group, Dollar Tree, eBay, Arthur J. Gallagher, Brown & Brown, Axon Enterprise, The Trade Desk, Airbnb, Boeing, Expedia, Target and Charles Schwab — companies expected to use AI rather than sell the hardware behind it.

Why are there no semiconductor stocks on the list?

The premise is that chip and infrastructure suppliers captured the first wave of AI economics because customers paid them up front for scarce capacity. The next wave, in this framing, accrues to companies that convert that spending into lower costs and better pricing inside their own operations, which shows up in operating margins rather than in vendor order books.

How did the named stocks trade on the day the list circulated?

Poorly. As of the last trade at 17:35 GMT on Aug. 17, 2026, all 11 of the named stocks with live quotes were lower. The Trade Desk fell 5.30% to 13.39, CoStar Group 3.27% to 31.32 and Expedia 3.06% to 322.51. Dollar Tree was nearly flat, down 0.05% at 129.33. Goldman Sachs shares themselves rose 0.66%.

What was the broader market doing at the time?

Mildly weaker across the board. The S&P 500 tracker SPY was at $773.47, down 0.37% from a prior close of $776.34. The Nasdaq 100 proxy QQQ was at $730.01, off 0.14%, and Dow tracker DIA stood at $533.88, down 0.54%. The declines in the named stocks were therefore steeper than the indexes in most cases.

How would an investor test whether the adopter thesis is working?

Look for measurable operating evidence rather than commentary: gross-margin improvement, revenue per employee, conversion rates or take rates rising over consecutive quarters, and reduced labor or markdown costs at retailers. If a company raises spending on AI without disclosing an expected return, the outlay is a cost rather than a proven advantage.

What are the main risks in a basket like this?

Two stand out. Efficiency gains can be competed away, especially in marketplaces and travel where rivals buy the same tools and pass savings to customers as lower prices. And the group is diversified across discount retail, property data, insurance broking, aerospace and online travel, so it carries consumer, real-estate, rate and aircraft-cycle risk simultaneously.

Sources

• Goldman Sachs Reveals AI Stocks To Buy List — GuruFocus