Investors are parsing two distinct anxieties around artificial intelligence at once, and the data available so far points in different directions for each.
Two separate fears
The debate breaks down into two separate propositions. Fear one: AI will destroy incumbent software and design platforms. Fear two: AI-related spending on hardware and infrastructure will collapse. Both concerns cannot be equally valid at the same time, and the early evidence is beginning to separate them.
Case study: Adobe and the missing 'SaaSpocalypse'
Adobe (ADBE) provides a clear example of how the first fear may be exaggerated. The stock was up +4.47% today to $263.50 (as of Sep 14, 3:12 PM EDT) after HSBC upgraded the company to Buy with a $308 target. The analyst’s assessment was direct: "We have yet to see any material impact from AI-powered competitors." Adobe reported 12.7% year-on-year revenue growth in Q2 FY2026, and current RPO rose 13.1% year-on-year.
Those figures highlight a common dynamic in enterprise software: deep workflow integration and habitual use make platforms sticky. Users often remain with tools that have become embedded in their processes, even when new AI alternatives appear.
Figma (FIG) reinforces the point. The share price was up +3.17% to $23.94 today, though it remains down -54.92% over the last 12 months. Despite the market’s earlier disruption fears, Figma’s operating picture shows 46% year-on-year revenue growth, 139% net dollar retention, and management raised full-year guidance to about $1.43B.
Fund managers in Australia are seeing a similar pattern. They noted that steep selloffs in names such as Xero and WiseTech, which have fallen 60-70% amid AI concerns, may present mispricing opportunities. Some observers draw parallels to historical episodes where markets overreacted to technology transitions.
Verdict on the first fear
Based on the evidence above, the danger that AI is immediately dismantling enterprise software appears almost certainly overblown. Network effects, workflow lock-in and compliance considerations are keeping many platform customers in place, and AI capabilities are being folded into existing systems rather than replacing them outright.
Where the second fear gains traction
NVIDIA (NVDA) illustrates the other side of the argument. The stock traded down -2.70% today to $212.39, and it was down -7.82% on the week (as of Sep 14, 3:12 PM EDT). NVIDIA remains up +11.86% year-to-date, but price momentum has clearly slowed after approaching a 52-week high of $236.54.
The more pertinent question here is not whether AI will disrupt NVIDIA, but whether AI capital expenditure will stay at current elevated levels. Hyperscaler spending is substantial, and markets are right to periodically test whether return on investment will justify continued high levels of capex ahead of each budget cycle.
Structural evolution of the AI trade
Another under-appreciated point is that the composition of the AI investment theme is shifting. The initial wave centered on compute, especially GPUs and companies such as NVIDIA. The emerging wave emphasizes coordination layers: orchestration, memory systems, networking and agentic infrastructure. Market pricing that still treats AI primarily as a GPU story may be overlooking a broadening architecture. That misalignment between perception and the evolving technical stack is a likely source of future mispricings.
The Honest Scorecard
| Fear | Verdict | Key Evidence |
|---|---|---|
| AI destroys SaaS/design software | Almost certainly overblown | Adobe 12.7% growth, zero disruption visible |
| AI capex cycle peaks too soon | Legitimate - watch carefully | NVDA -7.82% this week |
| Markets miss the agentic infrastructure wave | Underappreciated | CPU, DRAM, orchestration TAM expanding |
History shows that markets often fear disruption long before its practical effects arrive, while worries about bubbles can surface even as a structural buildout accelerates. Both anxieties create significant volatility and, for investors able to distinguish signal from noise, potential opportunity.