Deutsche Bank says its internal artificial intelligence platform is dramatically speeding up decisions on smaller corporate loans, compressing what once took weeks into minutes. The development comes as financial regulators worldwide sharpen their focus on the risks that AI can introduce into banking.
Marie-Jeanne Deverdun, the bank's technology chief, told Reuters that the platform, called ADA, went live over the summer and is already reshaping credit decisioning. Approvals for smaller companies can now move from weeks to minutes, and some decisions involving multinational borrowers have been cut from months to weeks.
What is ADA and how does it work?
ADA is not a single AI tool but an internal platform designed to build and run AI applications while keeping them under the bank's control. It includes monitoring, risk checks, and the flexibility to switch between different AI models as they evolve. That design is deliberate: by housing the technology in-house rather than relying on outside "black-box" vendors, Deutsche Bank aims to keep a clear line of sight into how decisions are made.
The pitch is straightforward. Faster, more standardized reviews can lift "throughput" — the number of loans a team can process in a given period — and cut manual work. In a business where costs are heavily tied to people, that kind of efficiency can matter a lot.
But the industry is learning that smarter software also raises the stakes on cybersecurity and governance. When AI systems touch sensitive customer and transaction data, the potential for misuse or error grows. Regulators are paying close attention, and banks that deploy AI without robust controls risk both fines and reputational damage.
Why this matters for investors
For Deutsche Bank shareholders, the "weeks to minutes" claim could eventually show up in a key financial metric: the cost-to-income ratio. If ADA genuinely reduces the staff time needed per loan decision, the bank could originate and service more smaller-company loans without adding headcount at the same pace. That kind of operating leverage typically improves the cost-to-income ratio, and it can make the bank more competitive in a segment where speed often wins business.
The catch is that AI only helps if it is controllable and auditable. By keeping ADA in-house, Deutsche Bank is betting it can satisfy supervisors' demands on model oversight, data handling, and cyber risk. That may require ongoing technology spending — the bank has budgeted more than €2.5 billion for technology between 2026 and 2028 — effectively making robust controls the price of admission.
Investors should watch whether the efficiency gains materialize without a corresponding rise in risk incidents. If ADA works as advertised, it could give Deutsche Bank an edge in the competitive market for smaller corporate loans. If it stumbles, the costs could outweigh the benefits.
Broader context: AI in banking
Deutsche Bank is not alone in exploring AI for lending. Across the industry, banks are testing ways to use machine learning to speed up credit decisions, detect fraud, and personalize services. But the technology also brings new challenges, particularly around explainability — the ability to understand why an AI made a particular decision — and compliance with fair-lending rules.
Regulators in Europe and elsewhere have been increasing their scrutiny of AI in finance. They want to ensure that models are not biased, that data is protected, and that banks can explain their decisions to customers and supervisors. The European Union's AI Act, which is being phased in, adds another layer of requirements for high-risk applications like credit scoring.
For everyday investors, the takeaway is that AI in banking is a double-edged sword. It can boost efficiency and profitability, but it also introduces new risks that banks must manage carefully. Companies that handle AI well may see improved margins and competitive advantages; those that mishandle it could face regulatory penalties and reputational harm.
As Deutsche Bank rolls out ADA more broadly, its progress will be a test case for whether AI can deliver on its promise in a heavily regulated industry. The bank's technology budget suggests it is committed to the effort, but the real proof will be in the results.
For more on how AI is shaping markets, see our coverage of AI earnings and central bank decisions and the ECB's upcoming rate decision.
