Fitch AI Stress Test Identifies Credit Risks Across Key Sectors

Artificial intelligence (AI) could trigger credit rating downgrades in several sectors under adverse conditions, although most issuers and transactions are expected to remain resilient, according to a new stress test by Fitch Ratings.

Fitch's Artificial Intelligence Stress Test evaluated credit exposure across corporate, financial institution, infrastructure, and structured finance sectors over a 5-year period. The analysis examined three adverse scenarios involving business disruption, over-investment, and asset impairment, each assigned an estimated 10–20 percent probability of occurring within 5 years. The scenarios are not mutually exclusive.

The assessment covered 107 subsectors across 21 broader sectors. Approximately 86 percent of subsectors received risk scores of 40 or below on a 100-point scale, indicating limited or no expected rating impact. However, 14 subsectors scored 60 or 80, suggesting greater downward pressure on credit factors that could result in downgrades of one or more notches.

The most vulnerable areas included services, data centers, and telecommunications, media, and technology. Insurance, real estate, transportation, and natural resources were among the least exposed sectors.

Under the disruption scenario, AI adoption could weaken established competitive advantages faster than companies can adjust their business models. Business process outsourcing and outsourced production services received the highest risk scores of 80. Insurance brokers, information technology (IT) services, and cybersecurity and IT operations scored 60, reflecting potential competitive pressures from AI-driven automation.

The over-investment scenario examined the consequences of weaker-than-expected returns on AI investments, reduced capital availability, and declining capital expenditures among major technology companies and other AI industry participants.

Semiconductors, memory and storage, and AI training facilities each scored 60. AI training centers in remote locations with concentrated, lower-quality tenant bases could face greater exposure. Graphics processing unit and AI compute asset-backed securities received a score of 50, the highest within structured finance, reflecting risks tied to AI-specific collateral and counterparty performance.

The asset-impairment scenario focused on deteriorating loans and investments involving AI-exposed businesses. Business development companies received a score of 40, indicating potential credit pressure. Private banking and wealth management also scored 40 in certain scenarios.

Most financial institution subsectors scored between zero and 20, supported by diversification, regulatory oversight, capital buffers, and established risk management practices.

Fitch said scores of 40 or higher could indicate negative rating implications for a representative entity under the applicable adverse scenario and rating criteria.