AI adoption may foster competition through several channels:
Labour substitution and augmentation: GenAI systems can automate cognitive tasks, particularly routine or repetitive tasks, lowering the skills threshold for market participation. Field experiments also suggest significant productivity gains, particularly for less experienced workers when they use AI to help perform higher-skilled tasks. This may reduce entry barriers in knowledge-intensive sectors by enabling smaller teams to perform functions that previously required larger, highly skilled workforces, supporting leaner and more agile business models.
Product improvement and innovation: AI can enable mass customisation and personalised services, supporting differentiation. Smaller firms can access design and communication tools previously reserved for larger incumbents. This expands the range of viable business models and opens space for challengers to compete on quality, user experience and new product features, rather than solely on scale.
Cost reduction, efficiency and productivity gains: AI adoption can reduce operating costs through automation, predictive analytics and modular deployment, which may lower fixed costs and support incremental scaling. At the same time, empirical studies report productivity gains in the form of time savings and quality improvements across a range of professional tasks. Taken together, these effects can enable leaner business models and facilitate entry into markets that might previously have been inaccessible.
Reduced search and switching costs: AI can also promote competition in downstream markets by reducing consumers’ search and verification costs. AI-enabled search, recommender and conversational systems help filter, rank and personalise options, which can broaden effective choice sets, improve matching efficiency and intensify competitive pressure on price and quality. Theoretical and empirical work shows that these systems save time and reduce effort, drawing in previously inactive consumers and enabling more efficient comparison across alternatives.
These benefits may not be evenly distributed as the competitive effects are highly context dependent. Adoption costs, integration challenges and access to enabling inputs such as data and compute may limit uptake, particularly among smaller firms. Moreover, some sectors remain less exposed to AI, and certain tasks may not be easily automated. Data access and the conditions under which firms can use or adapt AI models also influence competition both upstream and downstream. While concentrated control of data and cloud infrastructure may create advantages for large providers, widespread access to interoperable models and the ability to fine-tune them can support differentiation and entry. The competitive implications therefore depend on access terms, portability, and the practical ability of firms – particularly smaller ones – to tailor and integrate AI into their operations.
Furthermore, the pro-competitive effects of reduced search costs may depend on system design: if ranking or recommendation processes lack transparency or embed biases, they may channel demand towards particular suppliers and limit contestability. Ensuring fair visibility and portability across intermediaries is therefore central to realising these benefits.