Key Highlights

  • Core Scientific's proposed Merger with CoreWeave was terminated on October 30, 2025, after being widely misreported through much of that year.
  • Carlsmed is a genuinely AI-built medical-technology company, designing personalized spine-surgery solutions, despite sitting outside pure software.
  • Expensify reported second-quarter Revenue of about $33.9 million alongside a GAAP net loss of about $3.9 million as it migrates its platform.
  • Ziff Davis owns a portfolio of digital-media and Cybersecurity brands whose traffic Economics are directly threatened by AI-generated answers.

The tidy label "AI, software and technology services" is doing a lot of work in 2026. Underneath it sits a widening spread of Business models, all trading on the Nasdaq Composite yet pulling on very different levers.

First Advantage (NASDAQ:FA) (FA), Ziff Davis (NASDAQ:ZD) (ZD), Expensify (NASDAQ:EXFY) (EXFY), Carlsmed (NASDAQ:CARL) (CARL) and Core Scientific (NASDAQ:CORZ) (CORZ) are a useful cross-section. Two are core software and services companies. One is a digital-media operator wrestling with AI's disruption of the open web. One is a medical-technology name that happens to be built on AI. And one is an infrastructure story dressed up in the theme's clothing.

Grouping them is an analytical convenience, not a peer set. That distinction matters more this year than most, because the market is increasingly rewarding proof of AI monetisation over AI narrative.

The September 2026 backdrop

The macro frame is familiar but not settled. Rate expectations have driven the rotation between profitable, cash-generative software and longer-duration growth names all year, and quality-of-earnings has become the dividing line. Companies showing real free Cash Flow and Buybacks have generally been treated more kindly than those still spending toward a promise.

Within software and services, three sub-themes stand out. First, "applied AI", vendors embedding models into existing workflows (screening, expense management, content) rather than selling raw model access. Second, content and data licensing, where owners of proprietary datasets are negotiating, or litigating, with foundation-model developers. Third, the physical layer: the power, land and cooling that AI compute demands, which has pulled crypto-mining infrastructure squarely into the AI conversation.

Each of the five names below maps onto one of those sub-themes. None is a clean proxy for "AI stock," and treating them as interchangeable would be a mistake.

First Advantage (NASDAQ:FA): screening software with an AI story

First Advantage is the most straightforwardly "software and services" name here, an employment background-screening and compliance provider. According to aggregated figures, First Advantage reported second-quarter (calendar) 2026 results on 6 August 2026 with revenue of about $448.8 million, up roughly 14.9% year over year, and adjusted Earnings per share of about $0.35, both above consensus. The company raised full-year revenue guidance to roughly $1.69 billion at the midpoint and adjusted EPS to about $1.26, per the same source. Management, in the person of CEO Scott Staples, framed performance around an "AI-driven proprietary technology platform" and enterprise contract wins. First Advantage sits in the applied-AI camp, using automation to lower the cost of processing a very high Volume of background screens rather than selling AI directly. The reported improvement in Operating Margin (cited near 12.7% versus 9.7% a year earlier) is the more telling data point, because screening volumes are sensitive to hiring cycles.

If US hiring stays soft, volume could offset efficiency gains; the guidance raise suggests management does not expect that, but labour-market data will decide it.

Ziff Davis (NASDAQ:ZD): a digital-media veteran fighting the AI web

Ziff Davis is the theme's most conflicted case. It owns a portfolio of digital-media, martech and cybersecurity brands whose traffic economics AI-generated answers directly threaten. Per its earnings-call summary, Ziff Davis reported second-quarter 2026 results on 7 August 2026 with revenue of about $286.7 million, down roughly 2.7% year over year, and adjusted diluted EPS of about $1.03, up about 13.2% from $0.91. The same summary notes the company sold its connectivity business to Accenture for roughly $1.2 billion and repurchased about 4.5 million shares during 2026. Management indicated it is pursuing litigation with OpenAI and is wary of signing early content-licensing deals, preferring to hold out for what it calls fair compensation for Training data. Ziff Davis is effectively two stories stacked, a shrinking legacy-media base and an option on content-licensing and data value. The Accenture divestiture and aggressive buyback are classic capital-allocation moves to support per-share metrics while the Top Line declines.

The litigation outcome is genuinely unknowable; investors should treat any licensing upside as optionality, not a base case.

Expensify (NASDAQ:EXFY): a spend-management SaaS in transition

Expensify is a spend- and expense-management software company in the middle of a platform migration. According to its earnings-call summary, Expensify reported second-quarter 2026 results on 6 August 2026 with revenue of about $33.9 million, a GAAP net loss of about $3.9 million (improved from an $8.8 million loss a year earlier), non-GAAP Net Income of about $3.4 million and adjusted EBITDA of about $6.6 million. Average paid members were reported near 640,000 for the quarter. Free cash flow was about $6.4 million, and full-year free-cash-flow guidance was raised to roughly $12-14 million. The company said its "New Expensify" platform reached about $12 million in annual Recurring Revenue (up from roughly $7 million the prior quarter), with over 56% of users migrated, and it launched AI features including "agent rules," "custom agents" and an Expensify MCP integration for third-party AI assistants. The number that matters is the swing to positive free cash flow and non-GAAP profit against a still-declining or roughly flat top line, a familiar "profitability before growth returns" posture.

The migration and AI-agent features are early; whether they reaccelerate paid membership is the open question.

Carlsmed (NASDAQ:CARL): sector-adjacent, but genuinely AI-built

Carlsmed is deliberately included as a sector-adjacent name, and honesty requires flagging that clearly: it is a medical-technology company, classified under healthcare, not a pure software vendor. What earns it a place in an AI conversation is its product. Carlsmed designs AI-enabled, personalised spine-surgery solutions built around its "aprevo" platform, which pairs surgical-planning software and algorithms with custom implants. The company completed its IPO and began trading on Nasdaq on 23 July 2025. Trailing-twelve-month revenue was cited near $63.3 million (up about 65.4%), with a net loss around $36.4 million and full-year 2026 revenue guidance of roughly $74-78 million. Carlsmed is a reminder that "AI" now cuts across sectors. It carries med-device risks, reimbursement, adoption by surgeons, clinical evidence, that have nothing to do with the software names above, and it is early-stage and loss-making.

High percentage revenue growth is easier off a small base; the durability of that growth, and the path to profitability, are unproven. Treat it as a distinct, higher-risk holding rather than a software comparable.

Core Scientific (NASDAQ:CORZ): the merger that didn't happen

Core Scientific is the theme's infrastructure story, and its status must be stated precisely because it was widely misreported through 2025. Core Scientific and CoreWeave had agreed to a merger, but on 30 October 2025 Core Scientific announced the termination of that agreement after stockholders did not approve it at a special meeting, per the company's own investor-relations release. Core Scientific remains a publicly traded, independent company on Nasdaq under CORZ. There is no verified evidence of a revived or completed Acquisition as of this writing. Core Scientific has pivoted from bitcoin Mining toward high-density AI/HPC data-center capacity. Per a company release dated 27 April 2026, it plans to expand its Pecos, Texas campus toward roughly 1.5 GW of gross power (around 1.0 GW leasable), with a first data hall targeted for early 2027. CoreWeave, notably, has continued to exercise contract Options as a hosting customer despite the failed takeover. CORZ is not really a software stock, it is a power-and-real-estate play on AI compute demand, still carrying bitcoin-mining exposure and crypto-price sensitivity.

The leasing model's economics depend on customer concentration and financing; both are open risks.

What investors should watch

The unifying question across these five is monetisation proof, not AI presence. For First Advantage, watch hiring-volume data against the raised guidance. For Ziff Davis, the OpenAI litigation and any first licensing deal are the swing factors, alongside continued buybacks. For Expensify, the pace of New Expensify migration and whether AI agents lift paid membership will define the next few quarters.

Carlsmed sits apart: surgeon adoption, reimbursement decisions and evidence of a narrowing loss matter far more than any software metric. Core Scientific hinges on data-center lease-up, customer concentration (CoreWeave chief among them) and the cost of financing large builds.

Across all of them, rate expectations remain the shared macro variable, cash-generative names have been treated more gently than cash-consuming ones, and that regime shows no clear sign of reversing this month.

Monetization Proof, Not AI Presence

The label ties these companies together; the analysis pulls them apart. First Advantage and Expensify are the closest to a traditional software-and-services read, one scaling profitably and one turning the corner on cash flow. Ziff Davis is a value-and-optionality case shadowed by AI's disruption of web traffic. Carlsmed is an AI-built med-tech name that should be judged by healthcare standards. And Core Scientific is an infrastructure bet whose headline Takeover collapsed and whose future now rests on leasing power to AI.

Investors treating "AI, software and technology services" as one trade will misjudge at least four of these five. The theme is real; the homogeneity is not.