AI Project of the Year

Yes. Free end to end: no entry fee, no judging fee, and no charge for the badge, the certificate, the logo use, or any proof of winning. There is no cash prize either.

Anyone with an AI-enabled project that was genuinely delivered. You can nominate your own project (self-nomination), nominate a peer's project, or enter jointly (for example a vendor together with the practitioner team, with the spotlight on the practitioner leading adoption). The award is global from the first edition, and every entry must be beyond proof of concept: delivered, piloted, or operationalized, with evidence.

No, it is not a technology award, and every industry can enter. This award recognizes AI applied inside real project delivery, whatever the sector: construction and infrastructure, healthcare and life sciences, financial services, government and the public sector, manufacturing, energy and utilities, education, technology and software, telecommunications, retail and consumer, transport and logistics, agriculture, non-profit and humanitarian delivery, and others. That list is illustrative, not a menu.

The distinction that matters is not your industry, it is what AI did. Your project does not have to be an "AI project" in the sense of building AI as the product. A hospital rollout, a bridge programme, a benefits-system migration, or a factory changeover all qualify if AI was a meaningful part of how the work was delivered and the project was genuinely adopted. Edition one runs no sector tracks and no sector quotas, so every entry is judged against the same six criteria, and a well-evidenced public-sector project competes on exactly the same footing as a well-evidenced software one.

No. There is one submission funnel and one rubric. You enter once, for AI Project of the Year. From the finalist pool, the global jury designates Excellence in Responsible AI, and Excellence in Agentic AI only where a finalist qualifies. You never apply for the specials, and entering once puts you in the running for all three. Note the wording: up to three. Excellence in Agentic AI is only conferred where a finalist clears the bar, so an edition may confer two.

Two things at once. First, AI has to be a meaningful part of the project, not a decorative add-on, and the project has to have been genuinely delivered and adopted, not just launched. Second, the entry has to show responsible-AI practice: governance, risk, ethics, transparency, privacy, bias mitigation, and human oversight where applicable. The rubric rewards durable change and real adoption over model novelty.

Against a published six-criterion rubric (see the rubric), scored 1 to 10 per criterion. Judging runs in two tiers: a regional round across three macro-regions, then a global round in which finalists are re-scored from zero by the global jury. Only entries that clear a published cut-off advance, and the cut-off is a floor of quality, not a seat count. Judges are asked to distinguish claims from evidence, and to score the evidence.

The recognition package is a Credly-style digital badge, a digital certificate, logo and mark use, a place at the live online ceremony, and a finalist Story-of-Success publication in the community case-study library. Every finalist also receives a written, criteria-anchored feedback report of strengths and improvement areas. There is no cash prize.

Everyone gets an answer. Every entry submitted as final receives a written outcome notification, whatever the outcome. It tells you the furthest stage your entry reached and the date, and if your entry was ruled not eligible it tells you which rule it failed, so the reason is specific. It does not include scores, criterion marks, your rank, how close you were to the cut-off, who judged you, or jury comments, and it never compares your entry to anyone else's. The full criteria-anchored feedback report is a finalist benefit, so a notification is an answer, not feedback.

There is no appeal on judgment, and we would rather you knew that now. Scores, promotion at the cut-off, and the designation of the winner and the special recognitions are final. There is no formal appeal process and we do not enter into correspondence about a jury's judgment of a project.

There is one narrow exception, for facts rather than opinions. If something went wrong with the process, you may ask once, within 10 calendar days of your notification, for a factual-error review, on four checkable grounds: your entry was ruled ineligible on a factual mistake, it was assigned to the wrong macro-region, evidence you submitted before the deadline was not put in front of the panel, or an arithmetic error was made in applying the published weights. Disagreeing with a score is not one of the grounds. The review is decided by two organizers who were not involved and who have no conflict, you get the finding in writing either way, and where an error is confirmed we re-run that step correctly. Both the request and the finding go into the audit trail.

Judges are drawn from AIPM ambassadors and external experts, selected against published criteria and bound by published conflict-of-interest rules (a juror never judges their own entry, employer, client, or a connected project). Interest in judging is welcome, and the interest list has an option for it. Signing up expresses interest only; seats follow the published criteria and the conflict-of-interest rules, and no seat is promised.

Your project has to have been completed or operationalized on or after 1 January 2025, and on or before the nomination deadline of 23:59 UTC on Thursday 5 November 2026. Two published dates, no arithmetic. The opening date is fixed to the calendar rather than counted back from the deadline, so it does not move if the schedule does. That is the eligibility window for edition one, and it is applied strictly at both ends: there is no discretionary extension, no rounding, and no case-by-case exception for a project that falls just outside it, however close. We say this plainly because a window we bent for one entrant would not be a window, and because we could not apply it consistently across three regions. If your project falls outside, do not spend the effort on this edition; the next edition will have its own band and your dates may sit inside it. The complete eligibility rules are in the operating rules below, free to read now.

It is the 2026 edition. An edition is named for the year of the work it recognizes, not the year the ceremony falls in, which is how most annual awards work. So this is Edition 1 (2026): it recognizes AI-enabled project work from the 2026 cycle, and it is celebrated at the live online ceremony on 17 February 2027. Your badge and certificate will name that edition.

The interest list is open now. Nominations open on Tuesday 8 September 2026 and close at 23:59 UTC on Thursday 5 November 2026. Regional judging runs from 6 November to 11 December. Nothing is asked of anyone between 12 December and 10 January: that pause is in the calendar on purpose, because the jury is unpaid and the end of the year is not a period in which volunteer attention can honestly be requested. Global judging runs from 11 January to 5 February 2027, the live online ceremony is on Wednesday 17 February 2027, and feedback reports plus the Story-of-Success rollout follow through March 2027.

Edition one is settled on these points. It is judged by a jury, with no public vote. It is owner-funded, with no sponsor. It recognizes projects, and there is no separate person-level award in this edition. Eligibility runs from 1 January 2025 to the nomination deadline, as a fixed band rather than a rolling window. The full rules, rubric, and eligibility are in the operating rules below, free to read now.