AI scare selling looks overdone - here's what operators and the tape are saying
Fears that AI will crush margins and upend business models triggered broad selling across software, logistics, and security. Operators and analysts are pushing back. The common thread: AI is being integrated to drive growth and productivity, not wipe out incumbents.
Software: demand plus AI can expand, not compress
Figma (FIG) rose 6% Thursday after a 30% year-to-date slide, citing a record number of new customers in 2025 and faster revenue and product velocity. CEO Dylan Field put it bluntly: "As AI gets better, Figma gets better."
Figma Make - its AI tool for building web apps and interfaces with natural language - saw weekly active users jump more than 70% quarter over quarter. That momentum suggests AI features can widen the user base and use cases instead of cannibalizing them.
Logistics: incumbents are building, not blinking
Freight names like C.H. Robinson (CHRW) and Universal Logistics (ULH) sold off in double digits after a small Florida firm touted an AI platform promising 300%-400% volume scaling and a 70% cut in empty miles. CHRW has already bounced partway this week.
CEO Dave Bozeman was clear: "We are the disrupter and not the disrupted," noting the company's lead in AI integration. For diligence, watch how incumbents translate data scale into routing gains and load-matching efficiency - the unit economics will tell the story.
Cybersecurity: the pullback is creating setups
Wedbush's Dan Ives flagged buying opportunities in CrowdStrike (CRWD), Palo Alto Networks (PANW), and Zscaler (ZS), down roughly 7%, 16%, and 22% year to date. His thesis: "AI will be a major tailwind to the cyber security sector over the coming years as protection of use cases, data, and end points expand markedly."
Reality check from the engineering stack
Bernstein notes that coding is only about a sixth of an engineer's work. Research, testing, integration, reviews, documentation, and stakeholder alignment don't vanish with code assistants.
Their view: the de-rating of AI-exposed or adjacent tech looks overdone, and "over time, AI is likely to be absorbed into the toolbox of enterprise technology rather than replacing it entirely." That framing supports multiple expansion for firms that prove AI adds revenue while protecting margins.
What this means for portfolios
- Back incumbents with distribution, data, and AI shipping now. The mix matters: feature adoption, attach rates, and customer expansion beat slideware.
- Separate claims from economics. Track unit metrics: WAU growth for AI features, churn/retention, gross margin trajectory, and opex per dollar of ARR.
- Cybersecurity is structurally advantaged as AI expands the attack surface. Pullbacks in leaders can be entries when ARR growth and net retention hold.
- In logistics, model network effects and data moats. Look for sustained reductions in empty miles, automation per load, and yield per mile without margin erosion.
- Underwrite capex/opex shifts from AI inference and tooling. Build 2026-2028 scenarios with sensitivity to utilization, mix, and pricing power.
For deeper diligence on how AI affects valuation, risk, and forecasting, see AI for Finance.
Key data to watch next quarter
- Design/software: AI feature WAU growth, net revenue retention, paid conversion from AI workflows, gross margin mix.
- Logistics: loaded miles per truck, empty-mile percentage, automation per load, brokerage take rate, on-time performance.
- Cybersecurity: ARR growth, large-deal velocity, AI module adoption/attach, gross margin, net retention.
The market priced straight-line disruption. Operators are showing additive outcomes. If execution and metrics confirm it, this "AI scare" created mispricings worth leaning into - with eyes on the numbers, not the headlines.
Your membership also unlocks:
