As the costs of developing and maintaining advanced artificial intelligence systems rise sharply, market forces may drive AI toward natural monopoly or natural oligopoly. According to Professor Tejas Narechania of Berkeley Law this raises profound implications for competition, innovation, and public accountability. In this conversation, Professor Narechania explains how the infrastructure and computational demands of building advanced AI models create conditions that favor market consolidation and how regulation can be deployed to temper the risks associated, drawing comparisons to telecommunications law and other natural monopoly regimes.
The Risks of AI Market Consolidation
Professor Narechania identifies several critical legal concerns associated with increasing concentration in AI markets:
- Barriers to Entry and Competition: He explains how the high costs of developing state-of-the-art models can create barriers for smaller competitors, potentially raising antitrust concerns around market foreclosure and exclusionary practices.
- Bias and Accountability: Professor Narechania emphasizes that when a small group of firms controls foundational AI models, liability questions around bias, fairness, and accountability become more urgent.
- Data Access and Network Effects: He describes how dominant AI firms benefit from network effects, where access to massive datasets reinforces their market position. This raises legal questions about whether dominant players should be required to share data to prevent anticompetitive behavior.
- National Security and Systemic Risk: According to Professor Narechania, concentrated AI markets could also create vulnerabilities, with single points of failure or limited supply chains (e.g., chip production) posing national security risks.
Legal Tools and Responses
Professor Narechania outlines several possible legal interventions that could address the risks of monopolistic AI markets:
- Antitrust Enforcement: He discusses how antitrust tools might be applied to prevent market foreclosure and self-preferencing practices by dominant platforms.
- Interoperability Requirements: Drawing from telecommunications law, Professor Narechania explores whether legal mandates requiring interoperability between AI systems could promote a more competitive market.
- Public Infrastructure Models: He also explains how public utility frameworks could offer an alternative, with government-supported AI infrastructure ensuring broader market access.
A Legal Crossroads for AI Governance
Ultimately, Professor Narechania emphasizes that the law must decide whether to intervene early to prevent irreversible market concentration. While natural monopolies can offer efficiency in some contexts, he argues that proactive legal frameworks will be essential to ensuring fairness, innovation, and accountability in the age of AI.
Professor Tejas Narechania is a faculty member at Berkeley Law, specializing in telecommunications law, antitrust policy, and technology governance. His recent paper, An Antimonopoly Approach to Governing Artificial Intelligence, co-authored with Professor Ganesh Sitaraman, offers a detailed analysis of these pressing legal challenges.