Field Report

The AI Shakeout · Analysis

AI Was Supposed to Kill the Virtual Assistant. The Opposite Happened.

Inbox triage, scheduling, admin — exactly the work everyone assumed a chatbot would swallow first. So why is demand for virtual assistants still climbing in 2026? The honest answer flips the question most people are asking backwards.

A Kasb field report. Every figure below is sourced and linked at the foot of the page. It ends with a recommendation for one of our own tools — we'll be upfront about that when we get there.

Ask the internet whether you should still become a virtual assistant in 2026 and you'll get the same confident answer from a hundred directions: don't bother, AI's got it. On paper it sounds airtight. The core of the job — sorting inboxes, booking calls, chasing invoices, turning a founder's chaos into a tidy calendar — is exactly the kind of work a chatbot now does in seconds. If anything was getting automated first, surely it was this.

Except the data did something nobody puts in a doomscroll thread. It went the other way.

Upwork runs one of the largest freelance marketplaces on earth, and once a year it publishes what clients actually paid for — real completed-job earnings, not a survey of opinions. In its 2026 report, "general virtual assistance" didn't crater. It held its place as one of the most in-demand skills on the entire platform, steady year over year, even as AI tools spread into every corner of work.1

And the market behind those jobs is still compounding. The human virtual-assistant services industry — actual people, not Siri — sits at roughly $6.5 billion in 2026 and is on track to clear $40 billion by 2035, growing north of 20% a year.2

$6.5B
Human VA services market, 20262
~$43B
Projected by 20352
+109%
AI-related freelance work, YoY1
70%
More work finished, human + AI (up to)1

So the question "is being a VA dead?" has a clear answer. No. But it's the wrong question — and chasing the wrong question is exactly how people end up stuck. We'll get to the right one. First, what actually changed.

What actually changed

AI didn't replace the assistant. It changed the job.

Here's the part the headlines miss. AI didn't delete the work — it ate the boring half of it. The scheduling, the formatting, the first-draft email, the data entry: handed to automation, and good riddance. What's left is the half a machine still can't do well — reading a client who says "fine" and means anything but, catching the thing that wasn't in the brief, holding a relationship together, deciding what matters when everything's on fire at once.

Upwork's own economists put it plainly when they read the 2026 numbers: "AI isn't replacing people; it's sharpening where human expertise matters most."1 The same report found that pairing a person with AI lifted how much work actually got finished by up to 70% — even on simple tasks.1

— Upwork Research Institute, In-Demand Skills 2026

So the skilled assistant didn't get cheaper. She got faster. The same person now clears a workload that used to bury her, takes on more clients, and charges for judgment instead of typing. AI didn't end the job. It quietly promoted the people who were good at it.

That's why demand for people who can actually wield these tools is exploding — AI-related freelance work grew 109% in a single year.1 The market isn't paying for the absence of humans. It's paying for humans who bring the thing AI can't.

Every "AI killed this job" headline is really describing the same thing: the job got harder to fake, and more valuable to do well.

Now the part nobody tells you

Which doesn't mean you should be a VA.

This is where most "make money online" content quietly lies to you. It finds one path that's working, points a camera at it, and implies the path is the answer — for everyone, you included. It isn't. Virtual assistance is a real, durable, wide-open way to earn online. It is also not a default setting, and it is absolutely not for everyone.

Because being genuinely good at it asks for specific things. Honest question — is this you?

Reliable with no boss watching
You do the thing because it's due
Calm inside someone else's chaos
Their fire drill, your steady hand
Organised by instinct, not effort
Order is your default, not a struggle
Reads people who don't say it
You feel the "polite but unhappy" client

If that's you, VA could be one of the best doors you ever walk through. If it isn't, it'll feel exactly like every other thing you've tried — pushing a boulder uphill until you quit and conclude, one more time, that you're the problem.

You were never the problem. The problem was being handed someone else's door and told it was the only one.

Here's the truth the noise buries: there are far more ways to work online than the three or four that go viral — spread across every industry quietly growing, in corners nobody bothers to film. VA is one door on a very long corridor. Somewhere on it is a door shaped like you — where the skills you already have, from an ordinary job or an ordinary life, are an unfair advantage from day one.

The people who make it almost never had more talent or more grit than you. They picked a path that fit them — instead of the one with the best thumbnail. That's the whole game. Not hustle. Not a secret tool. Fit.

"Can't I just ask ChatGPT?"

A general AI will help. It won't get you all the way.

Fair — and since we've spent this whole piece on what AI can and can't do, let's be honest about this too. Yes, partly. If you've got a chatbot open, it's a real place to start thinking. Try these tonight; they'll beat another hour of YouTube:

Prompt 01 — the honest audit"Act as a blunt career strategist. My real skills are [X]. I can give [Y] hours a week. My budget to start is [Z]. I want my work to feel like [W]. List online paths that genuinely fit those constraints and rank them by how fast I'd realistically land a first client."
Prompt 02 — the reality check"For [a path, e.g. virtual assistant], give me the unromantic version: realistic beginner pay, time to a first client, true startup cost, and the top three reasons beginners quit."
Prompt 03 — the fit test"Here's the honest truth about me: [reliable / scattered / love admin / hate admin / great on calls / freeze on calls]. Is becoming a virtual assistant a good fit or a bad one for me — and what would fit me better?"

Use them. But know where they stop. A general model answers from internet averages, so it keeps recommending the same famous, crowded paths everyone already knows. It has no real scoring of how hard, how costly, or how fast-growing each path actually is. And it forgets every word about you the second you close the tab — a brilliant stranger, meeting you for the first time, every single time.

A decision this big — whether your next year is momentum or another dead end — deserves something built for exactly this one question.

Kasb

Here's the recommendation we promised

Kasb PathFinder finds the door that's shaped like you.

PathFinder is an AI built and trained for one job: this one. It reads you across eight dimensions — your real skills and assets, how your mind works, how you handle pressure, the way you communicate, your honest financial target, and what you actually want your life to look like — then scans broadly across growing industries (not a fixed list) to surface the paths where you hold a genuine unfair advantage, led by the single one that fits you best.

✓Your best-fit path — plus two strong alternatives, ranked by where your advantage is biggest
✓A personal profile summary, written for you
✓Your Vision Clarity & Execution Readiness scores
✓A personal SWOT — strengths, gaps, and openings
✓A clear-eyed read on where your edge actually is
✓An AI chat that actually remembers your context

You read this far because some part of you already knew the opportunity was real — and was tired of guessing which door is yours. This is where the guessing stops.

$29one-time payment · instant access
Show me where I fit →

No income promises and no guarantees — anyone selling you those is lying. What PathFinder gives you is one thing: a clear read on where your unfair advantage actually is, so the effort you were always willing to put in finally goes through the right door. Weigh that against another twelve months testing paths that were never shaped like you.

Stop guessing which door is yours. See where you actually fit.See my fit →

Sources & notes

  1. Upwork In-Demand Skills 2026 report: "general virtual assistance" remained one of the most in-demand skills year over year even as AI tools expanded; skills explicitly referencing AI grew 109% YoY; Upwork's Human+Agent Productivity Index found human–AI collaboration boosted project completion by up to 70%, even on simple tasks; economist quote re: AI not replacing people. Based on completed-job freelancer earnings, Jan–Dec 2025, U.S. marketplace. Upwork Inc. / Upwork Research Institute (Feb 2026).
  2. Human virtual-assistant services market estimated at roughly $5.6–6.5B in 2026, projected to reach ~$42–43B by 2035 at a ~22–23% CAGR (figures vary by methodology; estimates exclude AI chatbots/voice assistants, which are a separate, larger market). Business Research Insights, Virtual Assistant Market (2026); Wishup, 2026 Virtual Assistant Industry Report.

Kasb — Field Report · Figures reflect publicly reported data as of mid-2026 and may be revised as sources update.
Kasb Life is a brand of Zacreation LLC · 2105 Vista Oeste Nw Ste E, # 3096, Albuquerque, New Mexico 87120, USA.

We use cookies to improve your experience, analyze traffic, and measure our marketing. By continuing to use this site, you agree to our use of cookies. See our Privacy Policy.