Lloyds Banking Group has become the first company in Britain's FTSE 100 stock index to bring an AI assistant into its boardroom. The system, called a board bot, was built by a firm named Board Intelligence and lets directors ask questions in plain language about a board pack that can run to hundreds of pages. It flags inconsistencies, points out possible bias in a proposal, and pulls in outside information such as new regulations before a meeting even starts.
Right now the tool only helps directors prepare before they walk into the room. But the company that built it says the next step is letting it sit in on the meeting itself and speak up in real time, pointing out when someone might be repeating a mistake made before.
This is not actually the first time a company has let AI near a board seat. Back in 2014, a Hong Kong venture capital firm called Deep Knowledge Ventures gave an algorithm named Vital an actual board seat, with voting rights on investment decisions. That experiment stayed a curiosity for a decade. What has changed is that a major, regulated bank with millions of customers is now doing something similar, which makes it much harder for other big companies to ignore.
And it is not only banks. The chief executive of Logitech, a company known for computer mice and keyboards, has said her firm already uses AI in almost every meeting and that she would be open to a version of it formally joining her board one day.
The numbers explain why boards keep circling back to this topic. Research from Protiviti and BoardProspects found that only about one in four boards discuss AI at every single meeting. Among companies that report strong financial returns from their AI projects, closer to two in three keep AI on the agenda every time. A separate MIT study, cited by McKinsey, found that companies with boards confident in handling AI outperformed their peers on return on equity by close to eleven percentage points. Boards that ignore the topic are not just missing a trend, they are leaving measurable profit on the table.
None of this means a computer program is about to replace your board chair. Boards deal with questions about fairness, reputation, and long term consequences that a data model cannot weigh on its own. What is changing is the kind of person companies need in the boardroom: someone better at asking sharp questions and making judgment calls, and less valuable purely for knowing facts an AI can now summarize in seconds.
The practical risk sits in accountability. Under the EU AI Act and similar rules taking shape elsewhere, directors can be held personally responsible if they lean on an AI recommendation without being able to show they checked it properly. Any board, anywhere, that starts using these tools needs a clear, written answer to one question before its next meeting: which decisions can this system help with, and which ones must stay fully in human hands.