Geopolitics, cybersecurity, and regulatory divergence raise risks for global AI adoption – Moody’s

This press release was published in English using an automatic translation system

By Moody's

São Paulo, December 9, 2025 - Artificial intelligence (AI) continues to advance rapidly, as model performance leads to new discoveries and major providers expand deployment options for companies, according to Moody's latest report. However, concerns about a potential AI investment bubble are growing as investments in computing power and infrastructure far exceed the revenue generated by AI applications. At the same time, integration hurdles related to cost pressures and regulatory fragmentation create a more complex environment for adoption.

Intense competition is driving the rapid improvement of AI capabilities. Advances in leading AI models in the US have delivered significant progress in reasoning, multimodal capabilities and tool use for corporate integration. Open-source models, especially from China, are closing the gap with US proprietary systems, raising questions about the monetization prospects of leading AI providers.

The impact of AI across all sectors will expand, but value capture will become increasingly uneven. Although adoption is growing, productivity gains vary widely between sectors. AI continues to offer major benefits for routine, document-centered, or customer-facing tasks, but complex workflows still encounter friction. Even when models perform well in controlled environments, implementing AI in a company’s operations requires a complete redesign of processes. Consequently, we expect productivity gains to increase gradually but remain highly uneven, both across sectors and within each sector.

AI infrastructure is a critical bottleneck. Demand for computing power has driven a surge in the construction of data centers and long-term capacity commitments, while the shortage of specialized chips , network constraints, and rising energy needs are reshaping access to AI infrastructure. The consolidation of market share among a small number of cloud service providers is pushing prices up and widening the adoption gap between well-capitalized companies and peers with limited capacity to absorb the costs.

Geopolitics, cybersecurity, and regulatory divergence are amplifying AI risks. Geopolitical fragmentation is reshaping access to chips , computing, and data infrastructure, sometimes forcing multinationals to operate separate AI stacks in every region. The deeper integration of AI into workflows is increasing cyber risk and exposure to operational failures. Divergent regulatory regimes – from the European Union’s AI Act to China’s licensing framework  – will further increase compliance costs and complicate global deployment.