SerpApi.Org
The question now isn’t who’s building AI. It’s who’s actually deploying it.
The “Mag Seven” stocks have become the “Bag Seven” — or worse. That was fun while it lasted.
Through the first half of 2026, the group of megacap tech stocks known as the Magnificent Seven (eight if you count SpaceX ) were down from their highs. Meanwhile, the stalwart S&P 500 was up more than 9%. For the AI hyperscalers, “magnificent” has become mediocre and the trade has lost momentum.
It’s not hard to see why. Most of the AI-centered businesses are overpriced (roughly 7x revenue vs 2.7x for the S&P 500 ). Many have engaged in roundabout financing deals — Wall Street’s version of check kiting.
Since the obvious artificial-intelligence plays are overbought, investors have been left to look for alternatives. Those could be under-the-radar AI companies (already largely discovered), international stocks (currency risk with a strong U.S. dollar ), financials (credit risk) or defense (geopolitical risk).
For investors who believe in AI as transformational for business, there is one truly overlooked option: existing S&P 500 companies that have gone all in on AI.
Since February I’ve been researching more than 130 companies to gauge their adoption of AI, their investments and their business results — including whether they’ve been able to eliminate what’s often their biggest cost center: workers. The companies in this index were chosen deliberately — established large-, mid-, and small-cap businesses across 14 industries that use or could use AI, and not the Nvidias and Anthropics that sell it. The question wasn’t who’s building AI. It was who’s actually deploying it.
Companies talk a lot about AI, but just a handful have produced an impact that lands on the financial statement.
What I found for the vast majority of these companies is that in the three years since the launch of ChatGPT, most have talked a lot about AI, but just a handful have produced an impact that lands on the financial statement or regulatory disclosures. Most companies aren’t cutting jobs (though that number is growing). Most aren’t doing meaningful spending on AI. Most don’t have anything like a new, game-changing innovative tool (that works).
That said, some companies are genuinely using AI. For instance, Costco doesn’t disclose a lot about its use of AI, but analysts say the retailer is aggressively using it for inventory, pricing and supply-chain logistics. Maybe that’s true, but there’s not a lot of numbers coming out of the company. How much more efficient could it be? And I don’t trust Costco. You go in for a $1.50 hot dog and you walk out having spent $150.
Maybe there are a lot of companies like Costco out there that are hiding the ball when it comes to disclosing AI impact. But there’s a big difference between some nifty new AI inventory tracking and a real, bottom-line, game-changing shift. Investors need to know companies are not only adopting AI in significant ways, but also documenting it in financial statements, earnings calls, investments and workforce reduction.
So after analyzing three years of data, I found some strong AI adopters that were disclosing real impact. Then I asked one more question: Are the shares of the companies that are heavily committed to AI undervalued? I found quite a few, each with its own story.
Five companies stood out because of the depth of their documented commitment to AI — in filings, in capital allocation and in workforce decisions.
Of the 130 companies scored, roughly half showed meaningful AI activity by any measure — but even that overstates it. Most were at the awareness stage: AI is mentioned in a risk disclosure, nothing more. But five companies stood out not because they were obvious AI leaders, but because the depth of their documented commitment — in filings, in capital allocation, in workforce decisions — was genuinely surprising for companies most investors don’t think of as AI plays. These include a payments network, a century-old tech company and a fintech organization nobody’s heard of.
These five companies are doing the work:
1. Visa. The company with the most documented AI commitment in the entire dataset isn’t a tech company — it’s a 60-year-old payments network. Visa’s annual report quantifies 26,000 employees using AI tools, 261,000 AI-powered interactions and more than 100 internal applications built on the technology. The stock recently traded at 27x earnings, below the S&P 500 average. The catch (if there is one): Agentic AI threatens to boot Visa off the checkout page.
2. Salesforce. The software-as-a-service company that has most aggressively reorganized around AI rooted its entire product line in an AI platform called Agentforce, backed by more than $3 billion in acquisitions. Salesforce has cut thousands of jobs while simultaneously deploying AI across its customer-relationship management tools. The stock recently traded at 18x earnings. The catch: Salesforce’s moat is narrowing. AI-native competitors are building, more cheaply and quickly, what Salesforce sells. Going all in on AI is the right move. It may also be a race Salesforce can’t win.
3. ServiceNow. This enterprise-software company deployed more than $10 billion in AI acquisitions in a single year — including a $2.85 billion deal for Moveworks and a $7.75 billion deal for Armis. Its workflow-automation platform is being rebuilt around AI from the ground up. The catch: At a recent 67x earnings, you’re paying a steep premium for a transformation that hasn’t yet shown up in the financial results. The commitment is real. So is the execution risk.
4. Evolent Health: Evolent ranks ahead of every major insurer, hospital system and health-tech platform in the index. It uses AI to manage prior-authorization and clinical-decision support for health plans, which means it’s using AI to help insurers decide what care to approve and what to deny. That’s either the future of healthcare administration or a liability. The stock is down 39% over the past year, and its $654 million market cap is small enough that most institutional investors can’t touch it. The catch: Evolent’s business depends on health-plan relationships that can disappear quickly. And the phrase “AI-powered prior authorization” doesn’t sit well with many patients.
5. Pagaya Technologies: Pagaya is a fintech firm that uses AI to make consumer-lending decisions for banks and lenders — essentially an AI underwriting engine that sits between borrowers and the institutions that fund them. Its AI commitment is documented in its filings and its investment score in the 130-stock index is strong — it ranks ahead of Goldman Sachs and Netflix on documented AI adoption. Its market cap is about $1.4 billion, and its stock is down 45% over the past year. The catch: Pagaya operates in the subprime-adjacent lending market. When the economy turns, it turns fast — and AI doesn’t make credit risk disappear.
Ignore the AI noise and look for impact
Most AI companies are still figuring out what they’re selling.
The lesson from all five of these companies is the same one Costco has been teaching for decades: The best businesses aren’t the ones making the most noise. Instead, they’re quietly using every tool available to sell you something you need at a price you’ll pay. The difference is that Costco knows what it’s selling. Most AI companies are still figuring that out.
In the meantime, the companies that have gone all in on AI and can prove it — in their filings, in their workforce decisions, in their capital allocation — are trading at a discount to companies that haven’t done any of that. That’s your $1.50 hot dog. Whether it comes with a $150 cart of other stuff is up to you.
David Weidner writes about markets, money and the stories behind them. He is the founder of CapitAI, a Substack publication that publishes the CapitAI Index — a quarterly measure of AI adoption among publicly traded companies.
More: Can the ‘Magnificent Seven’ save a stock market that might be doomed without them?
Also read: The stock market has a ‘Magnificent Seven’ problem — but not the one bears are warning about
