Booking a flight to an AI conference is the easy part. Explaining the ROI to your boss when you get back is the real challenge.
If I were allocating your 2027 travel budget, I’d start by defining your primary goal: what to study, what to build, or who to meet. Each of these eight US AI events offers a distinct reason to step away from your desk.
1. AI DevWorld 2027
9-11 FEBRUARY 2027 | SANTA CLARA, USA
If you write software, start in Santa Clara. AI DevWorld, held alongside DeveloperWeek, covers AI engineering, agents, infrastructure and AI-assisted development.
I would bring an architecture sketch and ask where your application needs stronger evaluation, better tooling or simpler integration. Those questions give you a practical route through the sessions.
2. ACM TRUST 2027
7-9 MARCH 2027 | WASHINGTON, DC, USA & ONLINE
Who would trust your system with an important decision? ACM TRUST examines security, resilience, explainability, fairness and responsible computing.
I would attend when your team needs defensible evaluation criteria. Bring a concrete failure scenario and ask how researchers and practitioners would test it.
3. HumanX 2027
7-10 MARCH 2027 | LAS VEGAS, USA
HumanX is my pick when your next AI decision involves buyers, internal sponsors or business strategy. Its enterprise focus makes it relevant for leaders working through adoption and implementation.
I would choose one business process and investigate what successful change would require. For hospitality applications, my hotel technology events guide adds useful sector context.
4. NVIDIA GTC 2027
14-18 MARCH 2027 | SAN JOSE, USA & ONLINE
Your model's requirements eventually become infrastructure decisions. NVIDIA GTC brings accelerated computing, generative AI, agents and robotics into the same developer conference.
I would compare deployment options through memory, latency and compute demand. Take a representative workload so your questions stay specific.
5. ICLR 2027
26-30 APRIL 2027 | CALIFORNIA, USA
ICLR is a priority if your work depends on deep learning and representation learning. I would read relevant papers beforehand and arrive ready to discuss experimental choices and reproducibility.
California is announced, but the specific venue is still pending. Keep accommodation plans flexible until the organisers confirm it.
6. NAACL 2027
1-5 JUNE 2027 | SAN FRANCISCO, USA
Choose NAACL if language is the centre of your AI work. The conference covers natural language processing and computational linguistics, including language models, translation and evaluation.
I would bring examples involving ambiguous inputs or unreliable answers. Use them to ask which evaluation methods capture the problems your users encounter.
7. CVPR 2027
20-25 JUNE 2027 | SEATTLE, USA
CVPR is the research stop for computer vision and pattern recognition. Its workshops and tutorials run 20-21 June, followed by the main conference on 22-25 June.
I would organise your visit around one visual problem, such as recognition under changing conditions. Look for methods and collaborators that could improve your next experiment.
8. KDD 2027
1-5 AUGUST 2027 | SAN JOSE, USA
KDD connects data mining and machine learning with research and applied data science. I would investigate how teams connect datasets, experimental rigour and useful applications.
If you work on mobility, bring a question about prediction, demand or operations. My transport analytics events guide can help you find events where those methods meet transport decisions.
Which one belongs on your calendar?
If you are a PhD student, choose ICLR for deep learning, NAACL for language or CVPR for vision. I would send an applied data scientist to KDD.
Senior engineers should compare AI DevWorld with GTC, depending on whether application design or infrastructure is the bigger challenge. Business leaders should consider HumanX, while evaluation and governance teams should look at ACM TRUST.
TRUST and HumanX overlap in March, so choose deliberately. Before you book, write the one result that would make the trip worthwhile.