Some companies give AI to everyone. This CEO is doing the opposite

At London-headquartered payments technology company Paysend, access to artificial intelligence tools has to be earned.CEO Ben Chisell told CNBC he is a “very cynical man” about the hype around the technology, though he still sees value in it. 

Chisell, who previously worked at Amazon, eBay and Starling Bank, estimates that Paysend can now deliver around 25% more engineering work with the same-sized team, but argues that getting results isn’t as simple as handing everyone the latest tech.

Instead, he limits AI tools to his highest-performing employees, often only after they’ve proved what they can do with them. In the wrong hands, he said, AI can magnify bad habits such as laziness and poor judgment.

His broader rule is to resist the “fear of missing out” and focus on customer problems. 

“I know it sounds very unglamorous and unsexy as an approach given the world of tokenmaxxing and AI, but I think it’s really, really important,” Chisell said.

The pressure to demonstrate AI’s value is growing, from the legal sector to software engineering and finance, and many companies are responding by putting the technology in as many hands as possible.

London-based Lloyds Banking Group has rolled out 40,000 Microsoft 365 Copilot licenses, with 97% of employees who have one actively using it, the bank said. Newcastle-headquartered software company Sage said its AI coding tools had saved hundreds of thousands of work hours, while London-based professional services firm Grant Thornton UK has encouraged all of its staff to use Anthropic’s Claude as part of a £500 million ($660 million) investment in AI.

These figures do not necessarily translate into gains on the bottom line, however. U.K. government research published in January found that three-quarters of businesses using AI said it had improved workforce productivity, but 77% said revenue had not changed. 

Like Chisell, Lokdeep Singh, CEO of British enterprise software company Unily, sees AI as a “superpower” best reserved for high-performing employees. He is also wary of companies giving it out to everyone as a way to reduce costs.

“I personally believe you should use it to expand your ambition rather than make the team smaller,” he told CNBC. “But the reality is that in many industries, again, because of the macro environment, it is being thought as: ‘OK, can I cut costs? Can I cut employees?’ The truth is probably somewhere in between.”

At Paysend, Chisell wants employees to demonstrate what a tool actually contributes. New hires get a month to experiment with AI tools and explain what they have achieved.

He recently put that policy to the test with a new search engine optimization lead who requested access to Claude. A month later, the employee told Chisell he no longer needed it. All he found it useful for, he said, was producing briefs more quickly. So instead of a general-purpose AI tool, the employee suggested building a dedicated content-brief agent, which he is now working on.

At software company CoreView, the approach has been different. London-based CEO Simon Azzopardi told CNBC the company was initially “very gung-ho” on AI, initiating a widespread rollout for its software engineers. They first used AI to investigate problems by correlating logs, events and data, as well as to produce technical documentation and onboard new engineers faster.

Now, he said, developers increasingly coordinate teams of AI agents handling tasks including back-end and front-end coding, quality assurance and documentation. Azzopardi said work that previously took weeks could now take days.

However, AI-generated work isn’t necessarily productive if employees use it to build things outside their expertise, Azzopardi warned. “You feel more efficient, but you actually aren’t,” he said.

Still, when onboarding a major customer in the U.K.’s National Health Service, CoreView only needed to add one engineer while continuing to deliver its other projects, he said. The company has been hiring rather than cutting staff, using the extra capacity to take on more work.

As AI takes on more of people’s existing responsibilities, the question is what happens to the employees who used to do these tasks. Chisell believes companies will thrive if they help employees move into new roles, rather than replace them once their work is automated.

Customer-service workers, for example, have spent years resolving complex issues — a skill that’s valuable across a business, he said. For Chisell, technology should never replace human responsibility.

“Having a human in the loop shouldn’t mean human CC’d,” he said. “It should mean human ownership.”