Choosing an AI topic is harder than it used to be. The field moves so fast that a topic which felt cutting-edge two years ago can now read like history, and the ones that matter most in 2026, such as generative AI, AI agents and regulation, are still new to most classrooms and teams. The gap shows: in UNESCO's 2023 survey of just over 450 schools and universities, fewer than 10% had developed institutional policies or formal guidance on generative AI.
That is why this list has 70 topics in artificial intelligence, grouped by what you need them for: research, presentations, final-year projects, seminars, debates and essays. Each group is split into labelled sets of about five, with our pick at the end, so you can choose something with a solid body of sources behind it or something still being argued about.
Quick facts
At a glance4 quick answers
How many topics are on this list?
Answer70, across seven categories
Which topics are newest?
AnswerGenerative AI, large language models, AI agents, AI literacy and the EU AI Act
Which suit a debate?
AnswerThe 12 debate topics, each phrased as a yes-or-no motion
Can I run a topic as a live discussion?
AnswerYes, with a live poll or anonymous Q&A on AhaSlides
Artificial intelligence research topics
Our pick: topic 3, large language models. Everyone has used one, and why they sometimes state falsehoods so confidently is an open research question with plenty to cite.
Set 1: Models and medicine5 topics
1AI in medical diagnosis
What it coversHow accurately models read scans, pathology slides and patient records, and where they still need a clinician's judgement.
2AI in drug discovery
What it coversUsing machine learning to identify drug targets, screen candidate molecules and predict protein structures.
3Large language models
What it coversHow models trained on vast amounts of text generate language, and why they sometimes produce confident but false answers ("hallucinations").
4Transfer learning
What it coversReusing what a model learned on one task or domain to improve performance on another with less data.
5Fairness and bias in AI
What it coversMeasuring and reducing bias in the data and decisions of AI systems.
Set 2: Trust, safety and the planet5 topics
6Explainable AI
What it coversMethods that show why a model reached its output, so people can check and challenge it.
7Multimodal learning
What it coversModels that learn from text, images, audio and video together.
8AI safety and alignment
What it coversKeeping advanced AI systems doing what their designers and users intend.
9Energy use of AI
What it coversThe electricity and water cost of training and running large models, and how to reduce it.
10AI for climate science
What it coversUsing AI in weather forecasting, climate modelling and tracking emissions.

Artificial intelligence topics for presentations
Our pick: topic 20, AI agents. They are new enough that your audience will not have a settled view, and the risk of handing them control gives you a natural argument to build on.
Set 1: AI in everyday life5 topics

11Deepfakes and synthetic media
What it coversHow AI-generated video and audio are made, and how to spot them.

12AI in cybersecurity
What it coversDetecting phishing, malware and unusual network activity, and how attackers use AI too.

13AI in video games
What it coversHow game developers use AI to create believable characters and adaptive difficulty.

14AI for personalised learning
What it coversAdaptive tutoring systems that adjust content to each learner's pace.

15Smart cities
What it coversAI in traffic management, energy grids and waste collection.
Set 2: AI at work and online5 topics
16AI in social media
What it coversHow recommendation algorithms decide what each user sees.
17AI in marketing
What it coversAudience segmentation, personalised advertising and campaign optimisation.
18Data ownership and AI
What it coversWho owns the data that trains AI systems, and what that means for privacy.
19Generative AI at work
What it coversHow writing, coding and design tools are changing everyday jobs.
20AI agents
What it coversSystems that plan and carry out multi-step tasks on a user's behalf, and the risks of handing them control.
AI projects for the final year
Our pick: topic 22, the retrieval-augmented assistant. The scope is easy to control, the answers can be checked against the source documents, and the result is something you can demo live.
Set 1: Language and prediction4 topics
21Customer support chatbot
What it coversA chatbot that answers questions in one specific domain, such as a university's admissions office.
22Retrieval-augmented assistant
What it coversA question-answering tool that searches a set of documents before generating an answer, and cites where its answer came from.
23Emotion recognition
What it coversA model that classifies emotions from facial expressions or speech, with a section on its limits and ethical risks.
24Stock market prediction
What it coversA model that analyses historical market data, tested honestly against a simple baseline.
Set 2: Vision and city systems4 topics
25Traffic flow optimisation
What it coversUsing real-time traffic data to adjust signal timings in a simulated city.
26Virtual fashion stylist
What it coversOutfit recommendations based on a user's wardrobe and preferences.
27Fake news detector
What it coversA classifier that flags likely misinformation, and an analysis of where it gets things wrong.
28Plant disease identifier
What it coversAn image model that recognises crop diseases from photos of leaves.

Artificial intelligence seminar topics
Our pick: topic 36, AI and employment. Everyone in the room has a stake in it, so the discussion rarely stalls.
Set 1: AI in the real world5 topics
29How can AI help predict and manage natural disasters?
30How is AI used in medical diagnosis, treatment planning and patient care?
31What does responsible AI development look like in practice?
32How do self-driving cars perceive the road and make decisions?
33How is AI used in precision farming and crop monitoring?
Set 2: AI and society5 topics
34How can AI help detect and prevent cyber attacks?
35Can AI help address climate change, or does its energy use cancel out the benefit?
36How is AI changing employment and the skills people need?
37What ethical concerns arise from AI in autonomous weapons?
38What AI literacy do teachers and students need? (UNESCO's AI competency framework for teachers is a useful starting point.)
Artificial intelligence debate topics
Our pick: topic 50, students using generative AI for coursework. Almost everyone already has an opinion, and a show of hands usually splits the room.
Set 1: Minds, bias and rights6 topics
39Can AI ever truly understand or possess consciousness?
40Can AI decision-making ever be completely free of bias?
41Should facial recognition be allowed in public spaces?
42Can AI replicate human creativity and artistic expression?
43Will AI destroy more jobs than it creates?
44Should the companies that build AI systems be legally liable for their errors?
Set 2: Rules, risks and education6 topics
45Should AI-generated content always be labelled?
46Should there be a universal code of ethics for AI developers?
47Should governments strictly regulate the development of AI?
48Is artificial general intelligence (AGI) a realistic possibility in the next decade?
49Should AI be used in predictive policing?
50Should students be allowed to use generative AI for coursework?
For a debate on regulation (topic 47), the EU's AI Act is the obvious case study. According to the European Commission, it entered into force on 1 August 2024 and became applicable on 2 August 2026, with exceptions: prohibited practices and AI literacy obligations applied from 2 February 2025, while the rules for high-risk use cases in sensitive areas such as education and employment (Annex III) have been extended to 2 December 2027 under the AI Omnibus. UNESCO's Recommendation on the Ethics of Artificial Intelligence, adopted by all 193 Member States in November 2021, gives the global view to set against it.

Artificial intelligence essay topics
Our pick: topic 60, assessment in the age of generative AI. It is a live problem for every school and university, and a strong essay can argue for a concrete answer.
Set 1: Work, creativity and money5 topics
51AI and the future of work: which skills will matter most?
52AI and human creativity: collaborators or competitors?
53AI in agriculture and sustainable food production
54Opportunities and risks of AI in financial markets
55AI in mental health care: promise and ethical limits
Set 2: Care, trust and assessment5 topics
56Why explainable AI matters for public trust
57Robots in elderly care: an ethical assessment
58AI and cybersecurity: an arms race between attackers and defenders
59The privacy paradox: balancing AI innovation with data protection
60Assessment in the age of generative AI: how should schools and universities respond?

Interesting topics in artificial intelligence
Our pick: topic 70, who owns AI-made art or writing. The law is still catching up, so you can take a side and defend it.
Set 1: Ethics and law5 topics
61What are the ethical considerations of using AI in educational assessments?
62What biases could AI introduce into criminal sentencing?
63Should AI be allowed to influence voting decisions or election campaigns?
64Should AI decide who gets a loan?
65What are the challenges of combining AI with augmented and virtual reality?
Set 2: Technology and the wider world5 topics
66What are the challenges of deploying AI in developing countries?
67What are the risks and benefits of AI in healthcare?
68What are the limitations of current deep learning models?
69How can AI support wildlife conservation?
70Who owns a piece of art or writing created with AI?
Ways to run an AI debate or seminar with a group
AI topics work best when people get to argue about them rather than listen to one person summarise them. A few ways to set that up with AhaSlides:
- Poll. Put a debate motion to a live poll before you start, then again at the end, and see whether the arguments moved anyone.
- Word Cloud. A word cloud asking "What is the first word you think of when you hear 'AI'?" is a quick warm-up that shows how differently people see the topic.
- Q&A. People are often reluctant to ask "basic" questions about AI in front of colleagues, so an anonymous Q&A removes that barrier and lets the room upvote what it most wants answered.
- Open Ended. Ask "What worries you most about AI at work?" and read the replies aloud to start the discussion.
- Rating Scale. Use a rating scale for statements such as "AI will make my job better", then compare the average before and after the talk.
Tips for a discussion that gets somewhere
- Start from what people already know. The UNESCO survey above found that fewer than 10% of institutions had formal guidance, so for an AI-in-education seminar ask the room what their own school or team rules are before you debate what they should be.
- Vote twice. Poll the motion before and after the debate, so the result shows what changed minds.
- Keep sets short. Pick one set of about five topics per session rather than the whole list.
Frequently asked questions
What are the main types of artificial intelligence?
There are two common ways to group them: by capability (narrow, general and super AI) and by functionality (reactive machines, limited memory, theory of mind and self-aware AI). As IBM notes, narrow AI is the only type that exists today, and among the functional types only reactive machines and limited memory AI are real; the rest are theoretical.
What are the five big ideas in artificial intelligence?
Perception, representation and reasoning, learning, natural interaction and societal impact. They come from the AI4K12 initiative, jointly sponsored by AAAI and CSTA, whose wording is:
- Perception: computers perceive the world using sensors.
- Representation and reasoning: agents maintain representations of the world and use them for reasoning.
- Learning: computers can learn from data.
- Natural interaction: intelligent agents require many kinds of knowledge to interact naturally with humans.
- Societal impact: AI can impact society in both positive and negative ways.
What is a good AI topic for a beginner?
Recommendation algorithms on social media, AI in video games or deepfakes. Each has plenty of accessible sources and gives your audience something they have already experienced.
What is a good topic for an AI debate?
A yes-or-no motion people already have an opinion on, such as whether students should be allowed to use generative AI for coursework. Pick a motion with a real trade-off, so both sides have something to defend.
Related reading
- Controversial debate topics for more motions beyond AI.
- How to hold a student debate to structure the session.
- Interesting topics for discussion for a wider warm-up.








