ServiceNow and IBM earnings reignite AI fears, sending software shares lower

Enterprise software shares sold off broadly after ServiceNow and IBM results reignited AI-disruption fears, with ServiceNow down 13.1% and IBM down 7.1% in premarket trading. ServiceNow cited a 75 bps hit to subscription revenue growth from Middle East on-premises deal slippage, while Truist cut its price target to $120 from $125. The move also hit Salesforce (-4.5%), Oracle (-3%), Intuit (-2.9%), Adobe (-2.3%), Palo Alto Networks (-2.1%) and Palantir (-1.6%).

Analysis

This looks less like a one-off earnings reaction and more like a regime shift in how software gets valued: the market is repricing duration on the assumption that AI-native workflows will attack seat-based revenue before management teams can fully repackage products. The immediate losers are the vendors with the clearest “assistant over application” overlap, because investors now assume every AI feature launch from frontier labs becomes a pricing comparator, not just a marketing event. That is why the selloff is spreading beyond the names with direct misses — the multiple compression is being applied to the whole category as a proxy for future churn and slower net retention.

The second-order beneficiary is not necessarily another software vendor; it is the buyer of software, at least near term. Enterprises can use this volatility to force concessions on renewal terms, bundle commitments, and usage-based pricing, which pressures gross retention before it shows up in reported revenue. The real damage is likely to emerge over the next 2-3 quarters in billings growth and operating leverage, while headline revenue may still look superficially resilient because of contract lag.

The contrarian read is that the market may be over-penalizing the incumbents with deep workflow penetration and switching costs relative to pure horizontal tools. The names most at risk are those where AI can replace a narrow function quickly and cheaply; the least exposed are systems embedded in compliance, security, or mission-critical process orchestration. If the next earnings cycle shows even modest evidence that AI is augmenting ARPU rather than substituting for it, this selloff can reverse fast because positioning is clearly crowded to the bearish side.

Key Decisions for Investors

• Short the weakest AI-exposed horizontal software basket vs. the more embedded platform names: long PANW / short CRM or ADBE for 1-3 months, targeting multiple compression as investors differentiate between workflow-critical and feature-level software.
• Use the pullback to buy protective puts on NOW or CRM into the next 4-8 weeks if you own the sector, because the primary risk is not the current quarter but negative guide commentary on renewals and AI monetization.
• Avoid buying the dip in ORCL and IBM until there is evidence that AI is expanding spend rather than cannibalizing legacy software mix; if long, keep positions smaller and hedge with out-of-the-money puts through the next earnings season.
• For tactical traders, consider a short-term long vol trade in the software ETF or in NOW/CRM single-name options ahead of earnings, as the setup favors gap risk and implied volatility likely remains underpriced relative to earnings-day dispersion.
• If you want a contrarian long, wait for 2-3 days of stabilization and then accumulate the highest-quality workflow franchises with proven switching costs; risk/reward improves only after forced sellers clear and managements can rebut AI displacement with concrete pipeline data.