TL;DR: In 2025, Zapier began requiring a minimum level of AI fluency from every new hire, publishing a four-level rubric that ranks candidates from “Unacceptable” to “Transformative.” It’s the clearest sign yet of a shift Remote Work Europe has watched build all year: for remote workers and the self-employed alike, AI fluency has moved from a nice-to-have to a baseline expectation. This is what that means, and how fluent you actually need to be.
A year ago, “comfortable with AI tools” was a bonus line near the bottom of a remote job spec, somewhere below “good communicator” and “self-starter.” Today, at a growing number of remote-first companies, that bonus line is the spec.
That’s a big claim, so let me be precise about it, because the detail matters. There’s a real difference between an employer that requires you to be AI-fluent to get hired, and one that simply expects you to use AI once you’re in the door. Both are happening, and both change what remote work asks of you in 2026. Remote Work Europe tracks these shifts across the market precisely because our members feel them first: the people applying for remote roles, and the freelancers pitching for remote clients, notice the bar move before anyone writes a think-piece about it.
Here’s where the bar actually is.
What “AI fluency” means when a company puts it in writing
The most useful place to start is with a company that has spelled it out. Zapier, the remote-first automation platform, made AI fluency a formal hiring criterion in 2025, and published the rubric it uses to assess it.
Zapier’s framework has four levels. Unacceptable is the floor: what they describe as “a common starting point, especially for people who haven’t had the time, support, or context to explore AI meaningfully yet.” Capable means “you’re using AI on purpose.” Adoptive is where “AI is no longer an experiment; it’s a dependable part of how you work.” And Transformative is the top: “impact at scale. AI has fundamentally changed how you approach work.”
What’s striking is that a real company now scores candidates against this framework. In 2025 Zapier raised the bar again, so that applicants are expected to show AI embedded into their core work, with measurable impact, from day one. You don’t get points for having opinions about AI. You get assessed on what you’ve actually built with it.
The reason this example matters so much for remote workers specifically is that Zapier is fully distributed. This is a standard coming from exactly the kind of location-independent employer many of our readers want to work for, which makes it hard to dismiss as a big-office quirk.
Why remote employers moved from “nice to have” to “non-negotiable” on AI fluency
The Zapier policy is the sharp end of a much broader shift, and the clearest statement of the softer version came from Shopify.
In an internal memo made public in April 2025, CEO Tobi Lütke told his teams that “reflexive AI usage is now a baseline expectation at Shopify,” and set a test that has been quoted more than almost anything else in this debate: before asking for more people or budget, “teams must demonstrate why they cannot get what they want done using AI.”
Notice what that does. Rather than screening candidates at the door, the way Zapier’s rubric does, it reframes AI as the default tool of the job, the thing you reach for first, and puts the burden of proof on not using it. For anyone working remotely inside a company like that, fluency stops being a personal development goal and becomes the water you swim in.
This is the distinction worth holding onto, because sloppy coverage blurs it. “Required to be hired” (Zapier) and “expected to use it on the job” (Shopify) are different pressures. But they point the same way, and they compound: employers that expect reflexive AI use today tend to become employers that screen for it tomorrow.
Why clients now expect AI skills from freelancers
If you’re self-employed, you might reasonably think the memos of big employers don’t apply to you. The data says otherwise, and it’s blunt.
Upwork’s In-Demand Skills 2026 report found that demand for its top AI-applied skills grew 109% year over year – more than doubling. The growth wasn’t evenly spread, which is the useful part: AI integration demand rose 178%, AI data annotation and labelling 154%, and AI chatbot development 71%. As Upwork’s senior research manager Dr Gabby Burlacu put it, the data shows “rapidly growing demand for people who merge technical skill sets with more foundational ones.”
For a freelancer, this is the shape of the briefs arriving in your inbox, and the ones that never do. Clients who a year ago wanted a writer, a VA, or a designer increasingly want a writer, a VA, or a designer who works with AI, because they’ve worked out that the combination gets more done for the same fee. In the Connected job board, the remote roles crossing Diana’s desk each week increasingly assume AI is part of the toolkit rather than a specialism you’d hire separately.
AI literacy is the fastest-growing skill across Europe
Zoom out from any single company and the labour-market data lines up. LinkedIn’s Skills on the Rise 2025 analysis found that AI literacy is among the fastest-growing skills across regions and job functions, and named it the single fastest-growing skill in four of the eight markets it studied: the United States, Australia, Germany, and Spain.
Two of those four are major European economies, and the direction is the same right across the continent. Whether you’re working remotely from Berlin, Barcelona, Dublin, or a village in rural Portugal, AI literacy is a top rising skill in the market around you, showing up in the roles and the rates you’re competing for now.
The backlash against mandatory AI, and what it means for remote workers
It would be dishonest to present this as a tidy, one-way march, so here’s the counterweight.
When Duolingo declared itself “AI-first” in an April 2025 memo and talked about gradually reducing contractor work that AI could handle, the reaction was fierce enough that CEO Luis von Ahn publicly walked it back, later saying he had not given enough context and that he would not force employees to use AI. The trend didn’t reverse. The lesson is that mandating AI, clumsily, has real costs in trust and morale, and that the companies getting this right are the ones building fluency as a genuine capability rather than a loyalty test.
For you, as a remote worker or freelancer, that’s oddly reassuring. The bar is rising, but what wins is real, demonstrable fluency, the kind Zapier’s top two levels describe: using AI on purpose, then making it a dependable part of how you work.
How much AI fluency do you actually need for remote work?
Here’s the honest answer: further along than most people assume, but not as far as the hype implies. You do not need to become an engineer, or to have opinions about model architectures. You need to reach the level where AI is a dependable part of how you work, and where you can point to specific things you’ve done better, faster, or at all because of it.
A practical way to place yourself is to borrow Zapier’s ladder and ask which rung you’re honestly on. If you’ve experimented but nothing’s stuck, you’re at the start. If you reach for AI on purpose for real tasks, you’re capable. If you’d genuinely miss it if it vanished tomorrow, you’re adoptive, and that’s the level at which most remote employers and clients now expect you to operate.
If you’re a rung or two below that, the fix is deliberate, applied practice on the actual work you already do, not another twelve-month course, ideally with someone showing you the moves rather than leaving you to flail at a blank chat box. That’s the difference between knowing AI exists and being fluent in it, and in 2026 it’s increasingly the difference between the remote roles and clients you can win and the ones you can’t.
And if you’re a freelancer reading this with a sinking feeling that all of it is leaving you behind, let me be direct: you’re not too late, and you don’t need to become a technologist to catch up. Adding AI to years of real experience is far easier than starting from nothing. It’s part of why I’ve been building Solopreneur Superpowers with my colleague Jim, to help experienced people do exactly that, without the hype or the gatekeeping. If that’s you, come and find us.
Looking for remote roles that fit where you’re headed?
Connected is Remote Work Europe’s daily-curated job board for European residents: hand-picked remote roles, scam-filtered, delivered so you can see what employers are actually asking for right now, including how fast AI is moving into the requirements.
FAQ
Is AI fluency really required to get a remote job now? For a growing number of remote-first employers, yes. Zapier assesses every candidate against a published four-level AI-fluency rubric. Many others, like Shopify, don’t screen for it at hiring but expect reflexive AI use on the job, which amounts to the same requirement a little further down the line.
What counts as “AI fluency”? Using AI on purpose for real tasks, to the point where it’s a dependable part of how you work and you can show specific results from it. It does not mean coding, or technical AI knowledge.
Does this apply to freelancers, or only employees? Both. Upwork’s 2026 data shows demand for its top AI skills more than doubled year over year, so clients are increasingly expecting the freelancers they hire to work with AI, not around it.
Does this apply across Europe, or just the US? Across Europe. LinkedIn named AI literacy the single fastest-growing skill of 2025 in markets including Germany and Spain, and the same trend runs right across the continent, wherever in Europe you work remotely.
How do I build AI fluency without wasting months? Practise on the actual work you already do, rather than taking a generic course. Applied, guided practice moves you up the fluency ladder far faster than passive learning.