Finance firms adopt AI, but many struggle to show clear value

In short: Banks and other finance firms are using more AI, but many say it is hard to prove it is boosting profits and regulators are watching the risks.

What's going on

Finance companies are talking a lot about using artificial intelligence, but the real results look mixed so far. The Financial Times columnist Gillian Tett describes a New York finance firm that found its newest interns were comfortable using AI, but their ideas did not hold up when senior staff asked deeper questions. The firm reportedly made fewer return offers and is putting more focus on hiring for critical thinking skills.

New surveys show why opinions differ. A survey from Nvidia says 89 percent of finance executives believe AI is boosting revenue. But a separate survey from Cambridge Judge Business School, based on responses from 628 finance and AI companies and regulators, found many firms struggle to measure AI’s value. Only 40 percent reported any profit boost, and 43 percent saw no change.

The Cambridge survey also found AI is mostly used in “back-office” work, meaning internal tasks like paperwork, reporting, and support systems (the office behind the counter, not the customer-facing side). Many firms rely on outside AI models, and OpenAI was the most used provider in the survey.

What to watch

Regulators are increasingly focused on safety issues. These include cyber attacks, and “herding,” where many systems make similar trades at the same time (like a crowd all running to the same exit). Watch for more rules and more “sandbox” testing, like the UK Financial Conduct Authority offering a controlled space for firms to try AI tools with support.

Source: Financial Times