AI Cuts Transaction Costs but Creates New Market Failures as Data Center Power Use May Double by 2030
Updated
Updated · Forbes · Jun 8
AI Cuts Transaction Costs but Creates New Market Failures as Data Center Power Use May Double by 2030
1 articles · Updated · Forbes · Jun 8
Summary
AI is likely to make markets more efficient, but not frictionless, because it lowers search and bargaining costs while creating new failures and worsening old ones.
17% growth in data-center electricity demand in 2025 illustrates the tradeoff: AI can improve information and efficiency, yet shift environmental and infrastructure costs onto grids, water systems and communities.
$12.5 billion in reported U.S. fraud losses in 2024 shows how the same tools that explain products and contracts can also cheapen deception through fake testimonials, images, invoices and credentials.
Large-model economics may also entrench market power, as scarce chips, cloud capacity, data and distribution favor incumbents; digital creative markets already show concentrated rewards, with just 80 Spotify artists topping $10 million in 2025.
Regulators including the FSB and ECB warn that widespread use of similar AI models could increase herding, correlated returns and bubble risk, leaving markets faster and cheaper but not necessarily better priced.