Paysend CEO makes staff prove AI is useful before handing it out

Paysend gives general-purpose AI tools selectively, asking employees to show what they add rather than rolling them out to everyone. CEO Ben Chisell told CNBC that new hires can experiment for a month, while the company’s experience suggests a more focused tool can beat a subscription nobody needs.

Watch Desk analysis

What happened

Chisell says access at the payments technology company is generally reserved for high-performing employees who can demonstrate a tool’s value. A new search-engine-optimisation lead asked to use Claude, then told Chisell after a month that it had mainly helped produce briefs faster. The employee proposed building a dedicated content-brief agent instead, which is now in development.

The approach sits alongside very different experiments. CNBC reports that Lloyds Banking Group has rolled out 40,000 Microsoft 365 Copilot licences, while CoreView says its developers now coordinate AI agents across coding, quality assurance and documentation. CoreView CEO Simon Azzopardi cautioned that feeling more efficient does not always mean being more productive.

The wider numbers are a useful brake on the rollout race: UK government research cited by CNBC found that 75% of businesses using AI reported improved workforce productivity, while 77% reported no change in revenue.

Why it matters

A tool can save time on a task without improving the business result. Paysend’s trial-and-review approach asks employees to show the work an AI tool actually helps with, while the proposed content-brief agent points towards solving a specific workflow problem instead of handing out another general-purpose subscription.

That does not establish that selective access is better than broad deployment. It does make the decision less abstract: companies can compare what staff achieve, what remains awkward and whether a narrower tool would do the job better.

Our read

The useful lesson is not “keep AI away from most staff”. It is to treat access as a testable business decision, not a badge of modernity. A month-long trial followed by an honest account of what worked is refreshingly less theatrical than counting licences and calling it transformation.

What to watch

• Whether Paysend’s proposed content-brief agent is built and what work it improves.
• How companies measure gains beyond time saved or tools adopted.
• Whether selective trials help staff move into new roles as tasks change.

Discussion spark: Should companies make employees prove an AI tool’s value before granting access, or does selective access risk keeping useful tools away from the people who need them?

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.