Finance and cybersecurity firms lean on trust as AI tools spread

In short: As more office work gets AI help, firms in finance and cybersecurity are betting that trust and human oversight will still set them apart.

What's going on

Executives in finance and cyber defence are mapping out where new AI add-ons for Anthropic’s Claude could change daily work. These add-ons are often called “plug-ins”, which are like extra attachments that let an AI assistant connect to specific tools or data, similar to adding an extension to a web browser.

The goal is to speed up routine tasks, like summarising documents, drafting reports, or pulling information from internal systems. But leaders are also drawing clear lines around work that still needs people in the loop. In fields like banking and security, a small mistake can be expensive or dangerous, so companies want checks that go beyond “the AI said so.”

This is where “trust” comes in. For many white-collar businesses, trust means clients believing that advice is careful, decisions can be explained, and sensitive data is handled properly. In practice, that often means stricter review steps, clearer responsibility when something goes wrong, and limits on what AI tools can access.

What to watch

Watch for companies to publish tighter rules on where AI can be used, especially for customer-facing advice and security decisions. Another sign will be more spending on testing and monitoring AI outputs, so firms can show clients and regulators how they control risk. If AI plug-ins spread quickly, “trusted workflow” may become a selling point, like a restaurant kitchen that lets customers see how food is prepared.

Source: Financial Times