Best AI and Machine Learning Research Conferences in 2027

Rows of poster boards in a dark exhibition hall with a few glowing diagrams, lit in cool blue

AI and ML research conferences for 2027 with checked dates and venues: ICLR, ICML in Rio, AAAI in Montréal, IJCAI in Kyoto, KDD, AAMAS and NeurIPS.

AI / AI Education / AI conferences

The three machine learning research conferences that set the agenda in 2027 are ICLR in California (26–30 April), ICML in Rio de Janeiro (31 July to 6 August) and NeurIPS, which its organizers have placed in Europe for 2027 without naming a city or dates yet. Around them sit the broader AI meetings: AAAI-27 at the Palais des congrès de Montréal (16–23 February), IJCAI 2027 in Kyoto (7–13 August) and ACM KDD in San Jose (1–5 August). We checked every date on the organizer's own site on 19 November 2026, and each entry says what we found and what is still open.

Key takeaways

  1. ICLR (26–30 April) is the first of the big three in 2027 and the only one with confirmed dates in North America. ICML moves to Rio de Janeiro; NeurIPS goes to Europe, city and dates to be announced.
  2. Kyoto hosts two major meetings in August: IJCAI 2027 (7–13 August) and ACL 2027 (17–22 August, on our NLP list). A research team can cover both general AI and language research in one trip to Japan.
  3. Applied teams get more from KDD (San Jose, 1–5 August) than from the theory-heavy venues. Agent builders should look at AAMAS in Hanoi (3–7 May), the main research venue for multi-agent systems.
  4. MLSys and FAccT are not repeated here. They sit on our AI engineering and AI governance lists, and computer vision and NLP venues have their own pages.

Which AI and machine learning research conferences are worth attending in 2027?

Ranked by how much a practitioner or technology leader gets out of each one. "Checked" names the source behind each number and when we looked.

EventDatesWhere
AAAI-27 (AAAI Conference on Artificial Intelligence)16–23 Feb 2027Montréal, Canada
ICLR 2027 (International Conference on Learning Representations)26–30 Apr 2027California, USA
AAMAS 2027 (Autonomous Agents and Multiagent Systems)3–7 May 2027Hanoi, Vietnam
AISTATS 2027 (Artificial Intelligence and Statistics)3–6 May 2027Montréal, Canada
COLT 2027 (Conference on Learning Theory)28 Jun – 2 Jul 2027Tokyo, Japan
ICML 2027 (International Conference on Machine Learning)31 Jul – 6 Aug 2027Rio de Janeiro, Brazil
KDD 2027 (ACM SIGKDD Conference on Knowledge Discovery and Data Mining)1–5 Aug 2027San Jose, California
IJCAI 2027 (International Joint Conference on Artificial Intelligence)7–13 Aug 2027 (satellite in Hengqin 15–17 Aug)Kyoto, Japan
NeurIPS 2027 (Conference on Neural Information Processing Systems)TBA (2027 edition confirmed for Europe)Europe
  1. TBA (2027 edition confirmed for Europe) · Europe · city and venue TBA

    For: ML researchers, research engineers, and the recruiters and investors who follow them

    The largest machine learning research conference, and the main hiring market for research talent each year. The 2026 edition runs 6–12 December in Sydney with satellite meetings, and the foundation has committed 2027 to Europe and 2028 to the western United States. Teams that plan to submit or exhibit should watch the official site, because the organizers warn that predatory look-alike events use the NeurIPS name.

    Checked: Organizer site (neurips.cc Future Meetings), 19 Nov 2026. City and dates not yet published; start date is a sorting placeholder only.

  2. 31 Jul – 6 Aug 2027 · Rio de Janeiro, Brazil · venue TBA

    For: ML researchers and research-led product teams

    The 44th ICML and its first edition in South America. ICML leans more toward methods and theory than NeurIPS, and its tutorials and workshops are a good way for engineers to catch up on a sub-field in a day. Brazil also brings in a Latin American research community that rarely gets this much attention at a top venue.

    Checked: Organizer site (icml.cc), 19 Nov 2026. Venue not yet published.

  3. 26–30 Apr 2027 · California, USA · city and venue TBA

    For: Deep learning researchers and frontier-lab engineers

    The youngest of the big three and the one most closely tied to deep learning and foundation models, with all reviews public on OpenReview. Registration opens on 23 December 2026. The organizer lists only "California" so far, so hold travel bookings until the city is confirmed.

    Checked: Organizer site (iclr.cc), 19 Nov 2026. Some third-party calendars name San Francisco; the organizer does not yet.

  4. 16–23 Feb 2027 · Montréal, Canada · Palais des congrès de Montréal

    For: AI researchers across the whole field, from planning and reasoning to machine learning

    The 41st AAAI covers far more than deep learning: knowledge representation, search, planning, multi-agent systems and AI for social impact. The co-located IAAI track on deployed applications and the EAAI education track make it the most useful general AI meeting for people who need breadth rather than one specialty.

    Checked: Organizer site (aaai.org), 19 Nov 2026

  5. 7–13 Aug 2027 (satellite in Hengqin 15–17 Aug) · Kyoto, Japan · venue TBA; satellite event in Hengqin, China

    For: AI researchers, especially in symbolic AI, agents and AI for science

    The 36th IJCAI, one of the oldest AI conferences, with special tracks in 2027 on AI4Science, AI and health, AI and creativity, and AI for social good. It sits ten days before ACL 2027 in the same city, which makes Kyoto in August the most efficient research trip of the year for teams working on both reasoning and language.

    Checked: Organizer site (2027.ijcai.org and ijcai.org), 19 Nov 2026

  6. 1–5 Aug 2027 · San Jose, California · San Jose McEnery Convention Center

    For: Applied scientists, data science leads and ML platform teams

    The research conference closest to production. KDD's Applied Data Science track is written by people running recommendation, fraud, forecasting and search systems at scale, and 2027 adds an AI for Sciences track. In Silicon Valley it is also an easy recruiting stop for teams hiring applied ML talent.

    Checked: Organizer site (kdd2027.kdd.org), 19 Nov 2026

  7. 3–7 May 2027 · Hanoi, Vietnam · JW Marriott Hotel Hanoi

    For: Researchers and engineers building agents and multi-agent systems

    The 26th AAMAS, and the research venue with the longest record on agents, negotiation, coordination and game theory. With so much industry work now framed as agentic AI, its papers on how agents cooperate, compete and fail are worth reading for anyone designing multi-agent systems, even teams that cannot travel.

    Checked: Organizer site (AAMAS 2027, hosted by the University of Warwick), 19 Nov 2026

  8. 3–6 May 2027 · Montréal, Canada · venue TBA

    For: Statisticians and ML researchers working on probabilistic methods

    The 30th AISTATS is a smaller, single-track-feeling meeting where statistics and machine learning overlap: causal inference, Bayesian methods, uncertainty and experimentation. Data science teams that run experiments or need calibrated models often find more usable ideas here than at the larger venues. 2027 adds a dedicated workshop day for the second time.

    Checked: Organizer site (aistats.org), 19 Nov 2026

  9. 28 Jun – 2 Jul 2027 · Tokyo, Japan · venue TBA

    For: Theoretical ML researchers and PhD students

    The 40th COLT is the main venue for the mathematics of learning: generalization, optimization, online learning and the theory behind reinforcement learning. Few practitioners need to attend, but research leads use its proceedings to judge which claimed advances rest on solid ground. Paper submissions close on 29 January 2027.

    Checked: Organizer site (learningtheory.org), 19 Nov 2026

Which AI research conference should a practitioner attend?

Pick the venue that matches the work, then the city. Researchers and research engineers working on models go to one of the big three: ICLR for deep learning and foundation models, ICML for methods and theory, NeurIPS for the widest spread and the biggest hiring market. Applied scientists get more from KDD, where papers come with production numbers. Engineers building agent systems should read AAMAS even if they skip the trip; most of the multi-agent failure modes now showing up in industry were described there first. Our guide to agentic AI architecture covers the production side.

Breadth is the reason to choose AAAI or IJCAI. Both cover planning, reasoning, knowledge representation and AI for science alongside machine learning, which suits people who need to know what the whole field is doing. AISTATS and COLT are for specialists: probabilistic and causal methods at the first, learning theory at the second.

Engineering leaders who want systems talks rather than papers will find MLSys on our AI engineering conference list, and FAccT on the AI governance list.

How do research conferences differ from industry AI events?

Research conferences are built around peer-reviewed papers, posters and workshops, and their programs are set by program committees rather than sponsors. That makes them the most reliable place to see what will reach products in one to three years, and the worst place to compare vendors. Expo halls exist at NeurIPS, ICML and ICLR, but they serve recruiting more than sales.

They are also cheaper than most industry events and publish nearly everything. Proceedings for ICML and AISTATS appear in PMLR, ICLR reviews are public on OpenReview, and many talks are streamed. A team that cannot travel can still follow the work; the trip pays off mainly for people presenting, hiring or meeting specific authors. For tools that help read the literature faster, see our AI research tools guide.

Listed here, or should be?

If an entry is out of date or something belongs on this list, email [email protected]. Listings are editorial and free. A featured, labeled Sponsored slot is available on request.

Related questions

Where and when is NeurIPS 2027?

The NeurIPS organizers have confirmed only that the 2027 conference will be held in Europe. City and dates had not been published when we checked on 19 November 2026. The 2026 edition runs 6–12 December 2026 in Sydney, Australia, with satellite meetings, and 2028 is planned for the western United States.

Which is the most important machine learning conference in 2027?

NeurIPS, ICML and ICLR are the three most cited machine learning venues. In 2027 ICLR comes first (26–30 April in California), followed by ICML in Rio de Janeiro (31 July to 6 August) and NeurIPS in Europe later in the year.

Can I attend NeurIPS, ICML or ICLR without presenting a paper?

Yes. All three sell general registration to anyone, and industry attendees often go for the tutorials, workshops, expo and hiring. ICLR 2027 registration opens on 23 December 2026, according to its site.

Which AI research conferences are in Asia in 2027?

AAMAS 2027 in Hanoi (3–7 May), COLT 2027 in Tokyo (28 June to 2 July) and IJCAI 2027 in Kyoto (7–13 August, with a satellite in Hengqin, China). ACL 2027 follows in Kyoto from 17 to 22 August.

Explore More

Ready to Find the Right AI Tools?

Browse our data-driven rankings to find the best AI tools for your team.