M&T Bank Goes All-In on AI Copilots for 15,000 Workers
M&T Bank just made a serious move, rolling out AI tools to over 15,000 employees across its entire operation. The bank has been quietly upgrading its tech stack for years, but this latest push into generative AI copilots marks a real shift in strategy. If you’re wondering what’s behind the hype around What is AI, this is basically it — regular workers getting AI assistants that help them code faster, handle customer queries, and spot risks before they blow up.
The rollout isn’t some half-hearted pilot program. Copilots are now embedded across customer service desks, engineering teams, and risk management units, which tells you M&T is playing for keeps. Honestly, the whole thing adds up to one of the bigger enterprise AI deployments you’ll see in banking this year. And the numbers are legit — we’re talking thousands of employees actually using these tools every single day, not just poking around the interface.
One interesting detail: the bank had to deal with legacy infrastructure before any of this could work. They spent years cleaning up their data pipelines and IT architecture so the AI models would actually have decent things to work with. That’s why these AI Models are finally producing useful results instead of just hallucinating nonsense. The organization also had to figure out token budgets and usage limits, which got pretty wild when you consider how much computing power goes into running something like this for 15,000 people. AI Tokens are basically the hidden currency of all this, and managing costs at that scale is no joke.
- Bigger than most corporate pilots: Most banks are still testing AI with a handful of teams. M&T deployed to 15,000 people across multiple departments at once, which is seriously impressive at this stage of the industry.
- Real proof that the tech overhaul payoff exists: After years of fixing old systems, M&T shows that doing the boring groundwork pays off when you go all-in on AI later. Not every bank is willing to make that kind of commitment.
- Cost scaling is the next boss fight: Running these tools for tens of thousands of employees means token costs can spiral fast. How banks manage that going forward will separate the ones who stick with AI from the ones who walk away.