共有:
Career × Generative AI

For job seekers who skip generative AI,
the ¥10M salary line keeps moving away.

A new survey Mynavi Tenshoku published on July 30, 2026 finds generative-AI adoption among job changers at 46.7% overall. Narrow the sample to those earning ¥10 million (roughly $67,000) or more, and that figure jumps to 90.7%. Whether you "can use AI" has quietly become a line that splits income brackets.

AI Navigate Editorial2026.08.027 min read
GENERATIVE AI ADOPTION BY GROUP 46.7% All job changers 53.7% 20s 61.7% 30s ~30% 40s-50s 90.7% ¥10M+ earners
01
The Survey

The "just knowing how to use it" era ended, with numbers to prove it

What Mynavi's job-changer survey exposed

On July 30, 2026, Mynavi Tenshoku published its "Survey on Generative AI Use Among Job Changers," polling 700 full-time employees who had changed jobs within the past year (fieldwork ran April 7-10, 2026, via web questionnaire). According to reporting from Innovatopia and Japan HR News Online, 46.7% of respondents said they used generative AI during their job search. By age, adoption was 53.7% for people in their 20s, 61.7% for those in their 30s, and roughly 30% for people in their 40s and 50s.

On its own, that would be an unsurprising story about younger workers adopting new tools faster. But slice the same survey by income instead, and the picture changes. Among the 43 respondents earning ¥10 million or more per year, adoption reached 90.7%. Income turns out to correlate with AI adoption more strongly than age does.

As AI Navigate's own feature "AI-Native Job Market: Postings, Salaries, Interview Prep" already noted, simply claiming "I can use AI" stopped being a differentiator in Japan's 2026 job market. This Mynavi survey is the first to attach a concrete number to that qualitative sense that "using AI" alone is no longer enough.

02
By The Numbers

46.7% vs. 90.7% — the gap between two groups

46.7%
Adoption among all job changers (n=700)
90.7%
Adoption among ¥10M+ earners (n=43)
48.6%
Top use case: drafting emails/chat messages

Among the 327 respondents currently using generative AI, the most common use was "drafting emails and chat messages" at 48.6%, followed by "creating materials" at 38.2% and "gathering and summarizing information" at 37.0%. Separately, among the 389 respondents with any generative-AI experience, the top perceived benefits were "broadened my thinking" at 62.2%, "more mental headroom" at 62.0%, and "improved work efficiency" at 60.4%. That points to AI being used less as a pure time-saver and more as a thinking aid.

03
Who It Affects

Who this hits, and how

The effect isn't the same across roles

This number lands differently depending on who's reading it. For engineers and MLOps roles, the bar has already moved past "can you use an AI tool" toward "can you talk about shipping RAG or agent systems into production" — these roles were already living the 90.7%-style reality before this survey put a number on it. For business and PM roles, this is the first time a third-party survey has backed up the instinct that citing concrete AI-assisted output in an interview — materials produced, research turned around — actually matters. For people not currently job hunting, the effect is less urgent since it's not directly tied to your current paycheck — though a growing number of companies are starting to apply similar yardsticks to internal reviews and promotions, so it isn't entirely irrelevant either.

All job changers (n=700)¥10M+ earners (n=43)
46.7% adoption90.7% adoption
Adoption skews by age: 20s 53.7%, 30s 61.7%Adoption skews by income more than age
Top uses: drafting text, building materialsUse-case breakdown not disclosed for this segment

Saying "I can use AI" isn't a weapon anymore.
What counts is being able to say what you changed by using it.


04
What To Do Next

Three things to do this week

Turning the numbers into something personal

01

Inventory what you've actually done with it

Write down, concretely, where you used generative AI — drafting materials, research, message writing — and how much time it saved, in numbers. "I can use it" is weak; "it cut a 3-hour task to 20 minutes" is what makes a resume or work history convincing.

02

Build one interview-ready story

Take the survey's top-cited benefits — "broadened my thinking," "improved efficiency" — and turn them into a single, specific work episode rather than an abstract claim. A quantified outcome (hours saved, throughput increased) makes it land.

03

Weigh your target salary against your own adoption level

90.7% adoption among ¥10M+ earners is a correlation, not proof that using AI automatically raises your pay. But if you're targeting that income bracket, you don't want to be the one person missing a skill nine out of ten peers already assume you have.

05
Read The Caveats

Don't take this number at face value

A few caveats are worth keeping in mind. First, this is a correlation, not proof of causation. People who already earn ¥10 million or more likely have higher baseline tech literacy and information sensitivity — they may simply be people who use AI well because they're already the kind of people who use new tools well. Reading this as "use generative AI and you'll hit ¥10 million" overreaches what the data shows.

Second, watch the sample size: the survey covers 700 full-time employees who changed jobs in the past year, of whom only 43 earned ¥10 million or more. A 90.7% figure on a base of 43 people implies roughly 39 individuals — nowhere near the robustness of a large, industry-wide study. Third, the survey doesn't measure the quality of AI use. Someone who drafts emails with AI and someone who ships RAG or agent systems into production both count as "using generative AI" in this survey, but their market value from that usage is nowhere close to equal. The numbers show a trend — they aren't a blanket endorsement of AI use regardless of what it's actually used for.