Technology Spending in Finance Faces Return Pressure

The Return Question

Financial institutions have invested heavily in emerging technology, but the timing and form of payback have often been left vague. For investors, M&A dealmakers and corporate strategists, the task is to identify which projects will generate revenue or cut operating costs, where spending is essential to contain risk, and which technologies merit early investment despite distant returns.

AI's Growing Bill

The sums committed to artificial intelligence make returns harder to defer. GlobalData expects the global market to grow from $131bn in 2024 to $642bn by 2029, a 37.4% compound annual growth rate. Financial services spending is forecast to rise from just over $18bn to at least $87bn over the same period. Banks and insurers have tested chatbots, document summarisation, fraud detection, credit scoring and background automation. Agentic AI could offer a clearer route to returns by planning steps and completing workflows with limited human involvement. Onboarding, compliance checks and real-time risk scoring are likely early applications. The useful question is which deployments can be linked to a line in the profit-and-loss account.

Cybersecurity: Justifying Spend

Cybersecurity spending has a different justification: the losses it prevents. Celent estimates banks worldwide spent up to $32bn on cybersecurity in 2025. Annual economic losses from cyber breaches and attacks are estimated at about $500bn worldwide. Generative and agentic AI allow attackers to produce personalised messages at speed and scale, while deepfake tools make manipulated voices and images more convincing. AI is strengthening defence too. A Bank of England survey found that 75% of responding UK financial firms were already using AI. Some AI fraud systems have intercepted up to 92% of fraudulent activity before transaction approval.

The central question for investors is which technology deployments can be tied directly to profit-and-loss outcomes, rather than counting pilots launched.

Quantum's Two Timetables

Quantum computing calls for patience. Systems could improve arbitrage detection, derivatives pricing, credit scoring and calculations of economic capital. Quantinuum estimates that investment in quantum computing by banks and financial institutions could rise from $80m in 2022 to $19bn in 2032. Hardware remains the principal constraint, with commercial scale-up expected between 2030 and 2035. The security implications are more pressing: a sufficiently powerful quantum computer could break widely used public-key encryption, possibly as soon as 2029. Quantum key distribution offers an early application in secure communications, with commercial deployments beginning in late 2024.

AR Finds a Purpose in Insurance

Augmented reality had a less convincing reception in banking. Although the global market is forecast to grow from nearly $30bn in 2024 to more than $87bn by 2029, most banks and wealth managers have ended early initiatives. Insurers are finding uses tied to specific operational needs. Claims teams can use AR for remote inspections, while underwriters can visualise risks before damage occurs. Environmental risk offers a particularly useful application: showing a customer the potential consequences of an event can make underwriting decisions easier to understand. If smart glasses gain consumer acceptance, payments providers could use AR to reduce friction at the point of purchase.

The Investor's Lens

For prospective investors and acquirers, evidence of a defined job within an existing process will matter more than growth forecasts. Each technology presents a different return profile: AI must justify its growing bill, cybersecurity spending is justified by losses prevented, quantum requires patience alongside urgent security preparation, and AR succeeds when tied to specific operational needs.