UIST 2026 — the ACM Symposium on User Interface Software and Technology — runs 2–5 November 2026 at the Detroit Marriott at the Renaissance Center, 400 Renaissance Center, Detroit. Workshops are on Monday the 2nd; the main conference runs Tuesday to Thursday. Registration is open.
If you only read one thing on the programme, read the keynote.
The debate
“Direct Manipulation vs. Interface Agents: The Debate Continues” — Wednesday 4 November, 11:00–12:30, in the Columbus room. Pattie Maes and Ben Shneiderman, moderated by Cliff Lampe.
In the late 1990s these two articulated the two competing visions for how interactive systems should work. Maes argued for interface agents: systems that learn from users, personalise, reduce information overload, and act on a user’s behalf when appropriate. Shneiderman argued for direct manipulation: visible objects, rapid feedback, reversible actions, predictability, and a strong sense of user control.
The organisers’ own framing of how it turned out is worth quoting, because it is more honest than most retrospectives:
The practical outcome was not that one side “won.”
Direct manipulation stayed foundational for GUIs, touch, creative tools and dashboards. Maes’s agents arrived too, mostly under other names — personalisation, collaborative filtering, recommender systems, context-aware assistance, copilots. What the field absorbed was the constraint rather than the conclusion: automation is useful, but it needs transparency, controllability, calibrated trust, and graceful recovery from errors.
Which is precisely the live question in every creative tool shipping right now. Every generative feature in a DAW, a 3D package or a code editor is a bet on where the line sits between acting on visible objects and delegating to something that infers your goal. UIST is restarting the argument with both original parties in the room, at the moment the second option finally works well enough to be dangerous.
There is a second keynote too: Mark Ackerman on “A Tentative Mea Culpa: A Look Backwards to Help the Next Generation of CSCW Researchers.”
The conference published its own marking
UIST has done something this year that more conferences should do: it published detailed statistics about its own review process, before the conference, including the parts that do not flatter anyone.
- 1,259 complete submissions to the papers track — a record high and a 33% increase on UIST 2025
- 1,175 moved past desk rejection into full review, each receiving four reviews
- 1,316 external reviewers and 158 committee members produced 3,271 reviews and 1,120 meta reviews
- Overall, 21.4% of submitted papers were accepted
And then the part that is genuinely unusual to publish — how much each stage of the process actually changed outcomes. By pre-rebuttal score band, the share of papers whose score moved:
| Pre-rebuttal score | Score changed | Increased | Decreased | Accepted |
|---|---|---|---|---|
| ≤ 2.00 | 4.3% | 2.1% | 2.1% | 0% |
| 2.01–2.50 | 21.3% | 9.4% | 11.9% | 1.8% |
| 2.51–3.00 | 64.1% | 26.5% | 37.6% | 19.5% |
| 3.01–3.50 | 71.2% | 54.6% | 16.6% | 77.9% |
| > 3.50 | 54% | 50.8% | 3.2% | 98.4% |
| All | 36.9% | 20.5% | 16.4% | 21.4% |
Read the 2.51–3.00 row. Two thirds of papers in that band moved, and more went down than up — 37.6% decreased against 26.5% increased. The rebuttal is not a formality and it is not a one-way ratchet. Meanwhile a paper above 3.50 going in was accepted 98.4% of the time, and a paper at or below 2.00 was accepted zero times out of all of them.
The programme committee meeting also mattered: 15.2% of all papers had an area chair recommendation changed during it, most often resolving a “Discuss” into a decisive accept or reject.
If you have ever wondered whether rebuttals are worth writing, that table is the best public answer anyone has given.
Eleven workshops, and no position paper required
All eleven workshops run on Monday 2 November — five full-day (9:00–17:00) and six half-day (9:00–12:15 or 13:45–17:00). Topics range across flavour archives, agentic reimplementation, accessibility, haptics and writing assistants.
UIST deliberately runs these differently from most HCI venues, and says so:
organizers cannot require a position paper or any similar longer submission, and there is no expectation that you submit your own previously published work in order to take part.
Where space is limited a workshop may ask a few open-ended questions with short answers, and that is the limit of what participants are asked for. Joining a workshop needs no separate paper submission — you add it to your registration. For anyone who has been locked out of a workshop by a two-page position paper due six weeks early, that is a meaningfully lower door.
Ten of the eleven are at the conference venue. One full-day workshop — Revibing HCI: Agentic Reimplementation as Community Practice — and the Doctoral Symposium are at the University of Michigan Detroit Center, 3663 Woodward Ave., about 2.5 miles north. The free QLine streetcar runs up Woodward from Congress Street to MLK Boulevard / Mack Avenue every 15 minutes or less.
Practical notes
The Travel Scholarship has closed — it covered up to $1,000 toward travel, accommodation, meals and registration for students and early-career researchers, with a 15 September deadline and 22 September notifications. Worth diarising for 2027 if that is you: the eligibility criteria explicitly prioritise people from under-represented or marginalised groups in STEM, those based in regions or institutions with limited travel funding, and anyone attending UIST or an international HCI conference for the first time.
One scheduling oddity the organisers have handled well: the US midterm elections fall on Tuesday 3 November, the first day of the main conference. UIST has posted detailed guidance urging eligible attendees to vote early in person (in Michigan, 24 October to 1 November) or drop a mail ballot at an official box rather than post it. If you are travelling to Detroit and eligible to vote, sort it before you fly.
Why we are covering it
NewsNix covers UIST papers constantly — it is where a large share of the interface research we write about lands, including SplatXR’s Unity toolkit and, this week, a reach-remapping study for humanoid teleoperation from Rekimoto’s lab. The conference itself has not had a preview here before, and it should have. Eleven days out, with registration open and the programme published, is the time.
Sources
- UIST 2026 — ACM Symposium on User Interface Software and Technology
- UIST 2026 Keynotes
- UIST 2026 Announcements, including the full review statistics
- UIST 2026 Workshops
- UIST 2026 Program
- Direct Manipulation, 1988 University of Maryland UIS broadcast — Ben Shneiderman on YouTube
- UIST 2026 Teaser — ACM SIGCHI on YouTube