A four-day week at full pay through AI efficiency, and staff did not cheer (「給与維持の週休3日制」AI効率化で実現 しかし社員が喜ばず……なぜ？)
B · Useful, limited ITmedia NEWS interview with the CEO of SpiralAI on a four-day week introduced to force AI use; first-hand detail on staff reaction, but every outcome is the CEO's own impression and the company has no quantitative data yet.
Summary based on the full text·Original: about 7 minutes to read
In short
ITmedia NEWS (6 November 2026) interviews Yuichi Sasaki, CEO of SpiralAI, a Tokyo generative-AI start-up founded in March 2023. In June 2026 it added Friday as a day off for about 20 full-time staff without cutting pay, reducing weekly hours by about 20%, explicitly to make AI use mandatory rather than optional. Nobody welcomed the announcement: juniors looked anxious and a director challenged whether management could really take responsibility for deadlines. The routine settled in about a month, and the CEO says efficiency has risen by more than 20%, though there is no data.
Key numbers
- ~20%cut in scheduled weekly working hours from June 2026, at unchanged pay and daily hours
- ~20full-time staff covered by the four-day week
- ~3 months to ~1 monthtime to build a small app: a planner plus several engineers before, one planner with AI now
- 2 weeks / 1 monthperiod some staff were "afraid of Fridays" / time until the new cycle settled
As reported by the source.
Japanese·Web page·opens itmedia.co.jp
What it reports
- Scope: about 20 full-time staff, average age early thirties, no remote work. Daily hours unchanged; weekly scheduled hours down about 20%.
- Motive: the CEO saw in consulting that people resist changing how they work; cutting hours was chosen as a "drastic remedy" so that everyone must become 20% more efficient with AI.
- Reaction: "not a single person said 'great'". Staff worried about keeping client deadlines and quality; clients barely asked about the change.
- Some staff were "afraid of Fridays" for the first two weeks; after about a month work ran as before.
- Small applications that took a planner and several engineers about three months are now built by one planner with AI in about a month. In Slack, staff's AI agents sometimes talk to each other to add features or check specifications.
- No quantitative productivity data yet; no major problems with client deadlines or quality reported. The CEO's own sense is "well over 20%".
- The change is treated as one-way: reducing holidays again could be an unfavourable change to working conditions.
- The CEO thinks it suits firms with staff who have ideas to prove; where work is mainly following instructions, it may only add a day off. He says a 10-20% efficiency gain from uses such as AI-drafted email is possible in any industry.
How they got it
A single interview with the CEO by a freelance reporter, across three web pages.
Limits
- All outcomes come from the CEO; no staff interviews or output data.
- Small, young start-up whose product is AI; transfer to other firms is uncertain.
- About four months of experience at the time of the interview.
Why it matters for companies
Most firms leave AI gains to individuals, who often absorb or hide them. SpiralAI took the opposite route: cut the hours first and let the time constraint force the change. The useful detail is the reaction: staff worried about accountability for deadlines, not about losing a working day. Any firm trying this needs a plan for deadline ownership and should treat the change as hard to reverse.
Key passages in translation
Japanese original
みんな喜んでくれると思ったら、誰一人として『やった』と言わなかった
I thought everyone would be pleased, but not a single person said "great".
Japanese original
働く時間が20％減るので、全員が作業を20％効率化しないといけない。AIを使えたらいい、ではなく、使わないといけない状態にする狙いがあった
Working time falls by 20%, so everyone has to make their work 20% more efficient. The aim was to move from "it would be nice to use AI" to a state where you have to use it.
Japanese original
生産性の向上を裏付ける定量的なデータはまだないものの、取引先への納期やサービスの品質に、制度導入による大きな支障は出ていない。
There is no quantitative data yet to back up the productivity gain, but the system has not caused major problems for client deadlines or service quality.
Summary based on the full text. Read the original for the full argument and data.Read the original on itmedia.co.jp