Two reports of job cuts landed in my feed on the same day this week, with slightly different stories as to why. The companies don’t matter (but if you want to know, I’ve linked to the sources in the footnotes).
One, a SaaS company, recently told the SEC it’s cutting its workforce by roughly 20% and moving towards an AI-driven platform1.
The other, a UK utility, reported headcount reductions too – around 1,300 jobs over the next two years. This utility has spoken publicly about its AI transformation plans, but its chief executive was clear that this round is driven by customers’ changing habits (more than 90% now go digital first, and calls are down 20% year on year) and NOT by a desire to pivot to AI2.
A distinction without a difference?
The story a business tells about AI’s place in its tech stack is often not about the tech itself. It’s about who it’s trying to persuade, and the image it wants to project.
“AI-driven” reads as forward-looking, change-embracing, tech-first – attractive to markets and investors.
For staff and communities, it’s a less compelling story. Organisations with a public duty, an ethical standpoint, or values commitments that would conflict should be careful with their framing.
You should always be taking deliberate, documented decisions about AI use – and making sure your messaging matches them. Clear rationale matters when you’re facing supplier due diligence, data protection questions, or reputational risk from an AI error.
Looking for that AI boost…
The SaaSSoftware as a Service: instead of buying software outright and installing it on your own computer (like we did in the old days), you rent access to it online, often for a monthly or annual fee. The provider hosts it, updates it, and keeps the data — which is convenient, right up until you remember that also means they're the ones with their hands on the controls, not you, and they can cancel too. company went public in 2021. Its stock price dropped 21% in February 20263 on investor fears that agentic AIAI that can take actions toward a goal, browsing, running code, sending emails, rather than just answering a single prompt. Every tool with an API call gets called agentic now – before you use one, check what it's actually allowed to do without your say-so. would disrupt its whole model of SaaSSoftware as a Service: instead of buying software outright and installing it on your own computer (like we did in the old days), you rent access to it online, often for a monthly or annual fee. The provider hosts it, updates it, and keeps the data — which is convenient, right up until you remember that also means they're the ones with their hands on the controls, not you, and they can cancel too. work-management platforms.
So demonstrating that it’s responding to the challenge (or opportunity) of agentic AIAI that can take actions toward a goal, browsing, running code, sending emails, rather than just answering a single prompt. Every tool with an API call gets called agentic now – before you use one, check what it's actually allowed to do without your say-so. might reassure those investors.
That said, it might want to be careful – recent research suggests companies citing AI as a factor in job cuts have underperformed the Nasdaq by almost 10% in the 30 trading days following their announcement4, which implies the market doesn’t entirely buy the AI-restructuring story either.
…or brushing it under the carpet?
The UK utility is unionised, and the unions representing its workers have been raising concerns about chatbots replacing employees.
So the company’s framing – that this is about customers moving en masse to digital support – reframes the impact of AI.
Replacing people with AI purely for profit or efficiency wouldn’t be a good message. Replacing them to meet a customer demand that already exists could be something different.
“The typical attitude of tech executives has been to say that AI allows us to gain efficiency rather than admit that they overhired,” said Enrico Moretti, a professor of economics at UC Berkeley. “It’s an easy way out.”
US tech groups cut 140,000 jobs despite AI spending boom, FT (subscription required)
Where do you sit?
No matter your position on AI, there’s a story to tell, and how you tell it can affect your business.
It’s vital to carefully consider any AI adoption, with an impact assessment carried out. Lawyers are used to running compliance assessments – but the piece that gets missed, and shouldn’t be, is the impact on the broader business.
Does AI use fit your compliance obligations? Your ethics? Your position on developing local employment, or human connection? Do you want to be seen as tech-forward, or people-first?
Is AI already running as shadow IT among your staff? Or have you locked it down with strict controls and monitoring?
Any of these approaches could be the right one. The trick is making the decision deliberately, and owning it – so that when supplier due diligence, justifying your use of personal data, or an AI hallucination in the news comes calling, you know exactly why you took the decisions you took.
So what to do now?
OurSmall Business Health Check looks at six foundations that tend to matter most in your early days, when you’re scaling, or when you’re systemising or changing how you work — including, specifically, AI impact: what you’re already exposed to, and what to watch for if you’re considering using it more.
Book your small business health check
- https://d18rn0p25nwr6d.cloudfront.net/CIK-0001845338/8caa6f84-8fce-4956-b817-dd7ae68b450d.pdf ↩︎
- https://www.personneltoday.com/hr/centrica-to-axe-1300-jobs-in-next-two-years/ ↩︎
- https://www.forbes.com/sites/donmuir/2026/02/04/300-billion-evaporated-the-saaspocalypse-has-begun/ ↩︎
- https://www.ft.com/content/96a33881-27fd-42cf-8cff-4cbc87fc835f?syn-25a6b1a6=1 ↩︎