The AI finance revolution

ARTIFICIAL intelligence (AI) is steadily reshaping financial institutions, increasing operational risks and competitive pressures while offering new efficiency and investment opportunities, according to Fitch Ratings.

The agency’s latest sector review finds that AI adoption is creating both challenges and advantages across the industry, though immediate credit implications remain limited.

“Operational risk is set to rise across the board with the rollout of AI use cases along the value chain,” Fitch highlights, cautioning against pockets of heightened AI-driven credit risk in business development companies’ (BDCs) asset exposures and wealth management-related revenue pools.

AI is being used to improve fraud detection, customer service, underwriting, claims processing and data-driven decision-making.

Yet, material AI-driven cost savings have yet to materialise given use cases are often still in the proof-of-concept stage, Fitch notes, adding that efficiency gains are often reinvested to maintain competitiveness rather than to boost margins.

Within the financial sector, Fitch identifies BDCs as the most exposed to AI-driven disruption.

It notes that software accounts for an estimated 20% of rated BDCs’ portfolios at fair value, and while near-term asset quality deterioration is not expected, accelerating AI disruption in future years will be a challenge.

“Investor concerns about AI disruption risk for software companies have contributed to wider spreads for BDC bonds, higher redemptions and slower inflows for perpetually non-traded BDCs, and lower stock prices for publicly listed BDCs. This has weakened capital market access for BDCs,” Fitch points out.

“However, refinancing risk is manageable for rated BDCs given strong unsecured debt issuance over the past two years, which created additional capacity under secured credit facilities to fund unsecured debt maturities in 2026, if necessary,” it adds.

Wealth managers also face significant risks, as AI-enabled model portfolios and advice threaten traditional fee-based revenue.

“Wealth managers’ business models appear to be most at risk as AI-enabled model portfolios and advice can provide low-cost alternatives,” Fitch observes.

In contrast, alternative investment managers and insurers stand to benefit from AI-driven developments.

Fitch notes that some alternative investment managers have software exposure positioned to benefit from lower valuations or infrastructure investments, and that AI adoption may also accelerate privately funded large-scale infrastructure investment often backed by long-term hyperscaler leases.

Insurance companies can capitalise on rapid data centre expansion, with the agency stating that appetite and demand for coverage is exceeding insurance industry capacity.

However, growth carries risks including heightened catastrophe exposure, coverage uncertainty and new regulatory challenges.

Banks experience a mix of operational opportunities and indirect risks.

Direct credit exposure to AI-disrupted sectors is limited, and banks benefit from advisory and capital markets activity linked to AI investment.

Investment banks could, however, see revenue losses if market sentiment shifts or deals cannot be syndicated at expected terms.

Operational concerns

AI adoption is also raising operational and regulatory concerns across all segments in the financial sector.

Fitch describes it as “a double-edged sword”, with benefits accompanied by greater cyber threats, governance challenges, and potential model errors.

The agency warns that “flawed models would lead to inaccurate outputs, affecting decision-making, absent well-defined AI governance and a ‘human in the loop’”.

In insurance, AI-driven underwriting and claims processing could yield biased or discriminatory outcomes, and the report notes that “policy language was written prior to the introduction of generative AI, thereby increasing the risk of coverage disputes”.

The scale of AI adoption differs across institutions.

Large banks invest heavily in proprietary capabilities, regional banks combine in-house and third-party solutions, and community banks largely rely on core system providers embedding AI features.

For smaller institutions, AI may enhance efficiency, while for larger players, it can strengthen strategic decision-making.

Despite rising AI use, Fitch stresses that technology is not a near-term driver of credit ratings.

“AI is not a near-term rating driver for financial institution sectors. Adoption will be gradual and efficiency-focused, with credit outcomes still dominated by traditional considerations such as business conditions, financial flexibility and financial structure,” it explains.

Cost savings are unlikely to translate directly into higher profitability, as intense competition and rising servicing costs absorb potential gains.

“When substantial cost savings do occur, these will necessarily result in higher profitability margins. This is due to intense competition and rising AI servicing costs,” it explains.

It adds that this would be in line with prior technological advances, which did not result in structural improvements in efficiency ratios. Rather, AI investments will allow an institution to scale its business and remain competitive.

“AI adoption may give some financial institutions an early-mover advantage and efficiency gains, though competitors are likely to close the gap over time,” Fitch argues.

With AI, the financial industry is certainly in transition. Investors may need to reassess portfolio exposure across sectors, weighing potential disruption against emerging growth opportunities.

Allocations could shift toward firms best positioned to leverage AI for efficiency, innovation and competitive advantage.