NLP and Machine Learning Conferences 2027: ACL, IJCAI, AAAI

The research calendar in NLP and machine learning is crowded and confusingly named, and most lists of it are a wall of acronyms with dates someone typed a year ago. This guide pulls the dates from our AI events calendar, so the list is current, then explains who runs what, which conference suits which reader, and the honest answer to whether a business should send anyone at all.

Every NLP and machine learning research conference still to come in the calendar, in date order. Each name opens a page with the venue and a link to the organiser. Confirm dates and deadlines on the official site before you book.

If a conference you expected is missing, NeurIPS, ICML or EMNLP for example, it is because its next edition is not in our data yet. They run every year and will appear as soon as dates are confirmed. In the meantime, the organiser's site will have the call for papers, which is the date most researchers care about anyway.

The acronyms make more sense once you group them by the community behind them.

The ACL family. ACL is the annual meeting of the Association for Computational Linguistics and the flagship for NLP. Around it sit NAACL, the North American chapter, EACL, the European chapter, and EMNLP, which focuses on empirical methods. They share a reviewing pipeline, so for a researcher they function as a sequence of chances to submit through the year. For an attendee, they are the same community meeting in different places.

The machine learning big three. NeurIPS, ICML and ICLR. NeurIPS is the largest and meets at the end of the year. ICML meets in the summer. ICLR meets in the spring and has its roots in deep learning and representation learning. The communities overlap heavily: much of the same industry and academic crowd goes to all three.

The general AI conferences. AAAI and IJCAI. Older than the machine learning boom and broader in scope, with planning, reasoning, search, multi-agent systems and knowledge representation alongside learning. If your interest in agents is about how they plan and coordinate rather than how the underlying models are trained, these are often a better fit than the big three.

The applied data conferences. WSDM and KDD sit between research and industry: web search, recommendation, data mining. Smaller and more applied, with a higher share of papers from companies running systems at scale.

You publish in NLP. The ACL family is your calendar. Choose by deadline and by which venue your work fits best, and treat the regional chapters as real options rather than consolation prizes.

You publish in machine learning. The big three, chosen by deadline. Workshops at all three are often where newer or more applied work lands, and they are smaller rooms with better conversation.

You build language products and need to hire. ACL or the chapter conference nearest your team, plus NeurIPS if you can afford one large trip. Build a list of authors whose papers match your problems before you go, and book meetings.

You build agents. More of the relevant work than you would expect is at AAAI and IJCAI, and at the agent workshops of the big three. If your question is how to make an agent reliable in production, practitioner events are more useful still.

You are a student. Look for student volunteer programmes and travel grants on each conference's site. They exist at most of the major venues and they are the usual way students attend without a paper.

Many of the people searching for these conferences are not researchers. They run companies, or teams inside them, and they want to understand where AI is going. For most of them, the honest answer is no.

The talks are written for people who read papers. Sessions run to a tight format, the slides assume a technical background, and the interesting conversations in the corridor are about methods, not products. A head of marketing or operations will come home with a lanyard and very little they can use.

There are three exceptions. You are hiring researchers, in which case the conference is a recruiting event and worth every penny. You are selling to research teams, in which case the expo floor is your market. Or your product depends on a specific technique, in which case the workshop on that technique can be worth the whole trip.

If none of those applies, pick an event built for business readers. Our guide to AI conferences for marketing and sales leaders walks through the ones worth the time, and why. For a grounding in what agents actually do in a business, the explainer on what agentic AI is is a faster start than a week of research talks.

The strongest case is a paper on the programme with the company's name on it. Without that, write a one-page request that states four things.

If the budget will only stretch to one trip, pick the conference where the people you need are, not the most famous one. A chapter conference close to home, with the three authors you most want to meet, beats a flagship on another continent where you spend the week in keynotes.

Before: read the accepted papers list for your area and pick the authors you want to talk to. Message them a couple of weeks out with a specific reason. Researchers are generous with time at conferences, and most are glad to talk about their work with someone who has read it.

During: after each useful conversation, write one line about what you discussed and what you agreed to send. Poster sessions blur together quickly.

After: follow up within a few days with what you promised: the paper, the job description, the dataset question. The window in which someone remembers which of the hundreds of people they met you were is short, and it is the same window for a candidate as for a customer. Teams that exhibit and follow up at volume use Wysera for this: Wyse drafts each message from your note and waits for your approval before anything sends.

For business and regional events, browse the full AI events calendar or see the guide for event-led teams.