Where AI Competitive Moats Are Evaporating & Being Built

Artificial intelligence (AI) is dismantling the competitive advantages that entire market segments have relied on for decades. Understanding where competitive moats are disappearing versus where they are being reinforced due to AI may be one of the most important questions that investors can ask right now.Key Takeaways Traditional moats built on intentional friction, government lock-in, and information asymmetry are rapidly evaporating. New AI agents and data processing technologies are replacing them as they emerge. Fair Isaac Corporation’s (FICO) recent 26% stock plunge highlights the vulnerability of legacy monopolies to regulatory changes, as technology outpaces historical information asymmetries. Investors are finding new opportunities in thematic ETFs like (ROBO B) and (THNQ B-), which capture high-growth themes structural to the AI revolution. The Disappearing AI Competitive Moats in Enterprise SoftwareThe clearest example of evaporating moats is traditional enterprise software. Companies like Salesforce (CRM), which have long charged what amounts to a user experience tax for managing workflows and CRM systems, are facing a world where individuals can simply instruct an AI agent to build an equivalent system on their own hardware, Zeno Mercer, head of robotics and AI research at VettaFi, said. The friction and complexity that justified premium pricing is being commoditized from underneath. SaaS companies broadly are facing a similar reckoning, according to Mercer. A similar dynamic is playing out among legacy incumbents. Specifically, those with moats that relied heavily on information asymmetry, market friction, or government lock-in. Fair Isaac Corporation (FICO) spent decades enjoying a near-monopoly in credit scoring, holding a spot in the ROBO Global Artificial Intelligence ETF (THNQ) before being removed in June 2025. While FICO invested heavily in technology, advanced data processing tools made it easier to decipher financial datasets. Despite political scrutiny over anti-competitive practices and aggressive price hikes, FICO stock initially continued to appreciate. However, regulatory moats can vanish instantly. On September 29, FICO shares plunged 26% in its worst trading day on record. This followed the Federal Housing Finance Agency (FHFA) dismantling its 30-year mortgage monopoly by adopting the VantageScore 4.0 credit scoring model.Where New AI Moats Are Appearing for ETFsIf AI and regulatory shifts are dismantling legacy advantages, the physical world is where new moats are being built. In thematic investing, this shift highlights the concept of HALO (Heavy Assets, Low Obsolescence). Robotics and physical AI represent a “HALO²” investment theme (High growth and Immune/beneficiary of AI), according to Mercer. This offers both compounding growth potential and structural immunity from generative software disruption. Investors can access this theme through strategies like THNQ and the ROBO Global Robotics and Automation Index ETF (ROBO B), which focus on companies building the AI ecosystem. Companies like Advantech (2395.TW) are deeply embedded in the infrastructure of industrial PCs, robotics, drones, and rugged hardware. These form factors operate on manufacturing floors and in warehouses under conditions that software alone cannot replicate, Mercer said. Advantech packages sensors, actuation, and robot brains for OEMs and counts thousands of partnerships, including with Nvidia (NVDA). This deep integration gives it a market position that is exceptionally difficult to displace. Physical AI — turning electricity and motion into real-world tasks — represents a category where incumbents are primed to capture more value, not less, according to Mercer.Sovereign Compute and Security InfrastructureA second emerging moat is sovereign AI infrastructure. As enterprises grow nervous about feeding sensitive data into public platforms like OpenAI or Anthropic, demand is rising for private cloud solutions where companies run their own models on their own hardware through providers like Nebius (NBIS), Mercer said. A third moat is agent security and inference infrastructure. Companies like JFrog (FROG), which manage and secure the software deployment pipeline for AI systems, are positioned to capture growing GDP share precisely because they are building for the inference economy rather than the legacy SaaS world, Mercer said. Moats tied to complexity, artificial friction, and government-backed lock-in are being eroded by technology and policy. Meanwhile, moats tied to physical irreplaceability, proprietary infrastructure, and trusted security architecture are becoming more durable, according to Mercer. Looking for regular updates? Subscribe here for weekly insights on robotics, AI, and healthcare technology, delivered straight to your inbox. For more news, information, and analysis, visit the Artificial Intelligence Content Hub. vettafi.com is owned by VettaFi LLC (“VettaFi”). VettaFi is the index provider for THNQ and ROBO, for which it receives an index licensing fee. However, THNQ and ROBO are not issued, sponsored, endorsed, or sold by VettaFi. VettaFi and its affiliates have no obligation or liability in connection with the issuance, administration, marketing, or trading of THNQ and ROBO.