Overview
AI is spreading across firms, tasks, products, and start-ups, but access does not guarantee broadly shared competitive power. The series distinguishes users from developers and formal adoption from functional and task-level integration, then traces how skills, data, capital, workflows, innovation, market outcomes, and acquisitions shape conversion from an available tool into durable advantage. Early U.S., French, Portuguese, patent, and selected start-up evidence points both ways: adoption is uneven and some concentration mechanisms are visible, yet no general AI-driven markup surge or monopoly is established. The key test is whether complementary capabilities, imitation, entry, and independent growth distribute gains faster than ownership and critical inputs concentrate them.
Learning outcomes
- AI access can spread widely while the data, skills, workflows, and capital needed for effective use remain concentrated.
- Adoption is concentrated in larger and knowledge-intensive firms, while most early functional and task use remains narrow.
- Census evidence shows more task augmentation and capital change than AI-driven net employment change, with associations rather than causal proof.
- Early OECD user data show no systematic markup-growth increase, but capability and innovation patterns leave concentration risks open.
- AI start-ups attract entry and funding, yet acquisition can expand diffusion while reducing independent rivalry.
Keywords
Preview every chapter · Unlock full text, audio, and video from chapter 4
PlansFAQ
Does faster AI diffusion make markets more competitive?
Not necessarily: diffusion can widen entry and imitation, but concentrated skills, data, capital, or ownership can reinforce incumbents.
Why can AI adoption be widespread without equal competitive gains?
Because access to an AI tool differs from converting it into a reliable, scalable capability through skills, data, workflows, and investment.
What does the early evidence show about AI and market power?
It does not establish a general AI-driven markup surge or monopoly; it shows heterogeneous user outcomes, concentrated patenting in some periods, and acquisition-related concerns.
What should readers measure to judge AI's effect on competition?
Measure the actor and layer, adoption depth, denominator, time horizon, market outcome, entry, innovation, and acquisition, then separate correlation from causation.