AI can make business events more inclusive. It can also make exclusion harder to see.

As the business events industry embraces artificial intelligence, we need to ask a difficult question: are we using technology to remove barriers, or are we simply automating existing inequalities?

Artificial intelligence is quickly entering the business events industry.

It can help organisers create content, translate presentations, recommend sessions, personalise delegate journeys, analyse feedback and make information more accessible. It offers genuine opportunities to make events more efficient and, potentially, more inclusive.

But we need to have a conversation before we become carried away with what AI can do.

Technology is not automatically neutral. But efficiency is not the same thing as inclusion.

If we build AI into the systems we use to plan, market and deliver events without questioning the assumptions behind it, we run the risk of reproducing the very barriers our industry claims to be dismantling.

AI learns from a world that is already unequal.

AI systems are shaped by the data they are trained on, the decisions made by their developers and the way organisations choose to use them. Those inputs can reflect existing inequalities, stereotypes and gaps in representation.

UNESCO's Recommendation on the Ethics of Artificial Intelligence warns that AI can reproduce or amplify bias, possibly contributing to discrimination, exclusion and inequality. It also identifies human rights, diversity, inclusion and accountability as central considerations in the development and use of AI.

This matters because the world reflected in data is not always representative of the world we want to build.

Consider whose experiences are most visible in the material used to train an AI system. Which languages are well represented? Whose communication styles are treated as the norm? Which cultures, identities and lived experiences are missing or misunderstood?

These are not simply technical questions. They are questions about power, representation and whose experiences count.

And they're directly relevant to business events.

Where could the risks show up in our industry?

We don't need to imagine a distant future to see where the questions arise. They are already relevant to the decisions event professionals are making today.

1. Who gets seen, heard and recommended?

Imagine an association using AI to help identify potential speakers, analyse abstracts or recommend experts for a programme.

Used thoughtfully, these tools could help organisers discover people beyond their usual networks. Used carelessly, they could reinforce existing patterns of recognition and influence.

If the underlying data favours people who have historically been invited to speak, published more frequently or represented more prominently online, the system may continue to favour familiar names.

That could make an already narrow speaker pool seem like the result of objective analysis rather than the product of historical choices.

The same concern applies to recruitment, sponsorship prospecting and delegate recommendations.

AI might help us find a wider variety of voices. But it will not necessarily do so unless we deliberately look for them.

2. Accessibility is more than a feature

AI can offer real benefits for disabled delegates and people with different access requirements.

Automated captions, transcription, translation and conversational tools can help people access information in ways that work for them. They may also reduce practical barriers to participating in an international conference.

But these tools are not infallible.

Captions can misrepresent specialist terminology, names or accents. Translation can miss cultural context or alter the meaning of a speaker's words. AI-generated summaries may omit important information. A chatbot may provide an answer that is technically plausible but inaccessible, inaccurate or unhelpful to someone asking a question about their participation.

For someone relying on these tools to follow a session or understand essential event information, an error is not necessarily a minor inconvenience.

It can mean being excluded from the conversation.

The answer is not to reject AI-powered accessibility tools. It is to test them with the people who will use them, provide alternatives and ensure that there is a way to correct mistakes.

Accessibility should be measured by whether people can participate, not by how many AI features appear on an event technology checklist.

3. Personalisation can come at a cost

Many events collect information about their delegates to improve the experience. AI can make it easier to analyse that information and tailor communications, recommendations and services.

But how much do delegates understand about the information being collected and how it is being used?

Information about dietary requirements, accessibility needs, language preferences, or participation patterns can help deliver a better experience. Other data may be more sensitive, particularly when it relates to someone's health, identity or personal circumstances.

Organisers need to think carefully about what they collect, why they need it, who can access it and whether it is being shared with third-party AI providers.

There is also a wider question about inference. A system may attempt to conclude a person's identity from information they never intended to disclose for that purpose.

A delegate should not have to surrender unnecessary personal information to participate in an event.

Personalisation should improve people's experience without compromising their privacy or their ability to make choices about themselves.

4. Whose version of the world does AI present?

International business events bring together people with different cultural backgrounds, languages, political circumstances and experiences.

AI-generated content can help organisers communicate across those differences. But it can also flatten them.

A system may misunderstand a cultural reference, reproduce a stereotype or present one perspective as though it were universal. Generated images and marketing copy may look diverse while relying on superficial representations of people and communities.

Representing diversity is different from understanding it.

If we use AI to create event campaigns, educational materials or content about communities whose experiences are unfamiliar to us, human review matters. So does involving people with relevant knowledge and lived experience.

The goal should not be to make content look inclusive. It should be to make sure it is accurate, respectful and meaningful to the people it represents.

The problem is not just the technology. It's who is accountable.

When a human makes a decision that excludes someone, we can ask why that decision was made and who was responsible.

AI can make those questions harder to answer. Decisions may be influenced by systems that are difficult for users to understand, while responsibility becomes dispersed between software providers, event organisers and the people operating the tools.

But introducing AI does not make responsibility disappear.

The US National Institute of Standards and Technology's AI Risk Management Framework delivers a practical method to identifying and managing risks throughout the lifecycle of an AI system, including its use and evaluation. UNESCO's ethical framework similarly emphasises human monitoring, transparency, fairness and accountability.

For business events, these principles should translate into practical questions.

Before adopting an AI tool, organisations should understand what it does, what data it uses, where its limitations lie and what happens when it produces a harmful or inaccurate result.

They should also know who is responsible for reviewing its outputs, responding to complaints and correcting mistakes.

A vendor's assurance that a product is innovative, responsible or ethical is not a substitute for asking these questions ourselves.

Five questions every event organiser should ask

We don't need to become AI engineers to take responsibility for how we use the technology. But we do need to ask better questions.

1. Who might this tool disadvantage?

Test how it performs for people with different disabilities, accents, language needs and cultural backgrounds. Don't assume that good results for one group mean it works equally well for everyone.

2. What data is it using?

Understand what information is collected, how it is processed, whether it is shared with third parties and whether the use is necessary and proportionate.

3. Can people challenge or correct the outcome?

If an AI tool gives incorrect information, excludes someone from a process, or makes an inappropriate recommendation, there should be a clear route to human support and correction.

4. Have the people affected been involved?

Consult disabled delegates, neurodivergent people, LGBTQIA2+ professionals, speakers and other relevant communities when assessing tools that may affect their participation. Involve people in testing, not just in the final evaluation.

5. Who remains accountable?

Identify a person or team responsible for the tool's use, monitoring and consequences. Human monitoring must mean more than someone occasionally checking whether the system appears to be working.

Consider these questions before procurement, during testing, and throughout the event lifecycle, not only after something goes wrong.

The question we should be asking about AI in business events

The business events industry has a genuine opportunity to use AI to widen participation, improve accessibility and make international connections easier.

But we should be honest about the risks as well as the opportunities.

If we use AI to speed up existing processes without examining whether those processes are fair, we may reproduce exclusion more efficiently. If we use it to represent communities without listening to them, we may amplify stereotypes rather than challenge them. And if we cannot explain who is accountable when a tool causes harm, we have a governance problem, not just a technology problem.

We talk about the future of events as though innovation is inherently positive. It isn't. Innovation only matters if we are thoughtful about what it changes, who benefits and who bears the cost.

The real test of AI in business events is whether it helps more people participate, contribute and belong, without creating new barriers along the way.

Before asking what AI can do for our events, perhaps we should ask a more important question:

Are we building a future that includes more people, or simply making the systems we already have more efficient?

---

This article draws on UNESCO's Recommendation on the Ethics of Artificial Intelligence and the US National Institute of Standards and Technology's AI Risk Management Framework. These provide useful foundations for considering human rights, fairness, transparency and accountability in the use of AI.

---

The uncomfortable conversations don't stop here.

New Intent challenges the ideas, decisions and systems shaping the business events industry.

Get our latest articles, thought-provoking perspectives, research and practical resources on human rights, inclusion and accountability, delivered straight to your inbox.

Independent thinking. Difficult questions. Ideas for change. Sign up now for the New Intent Briefing