Nathalia Vega, research director at innovation consultancy Designit, says the sky is the limit for airports is they learn from the successes and failure of AI

It’s predicted that the market for global AI in aviation is set to soar to $4.04 billion by 2032.

The implementation of AI in airports is bound to come with its own set of triumphs and challenges – and some will fare better than others.

It comes down to how AI is being integrated into airports worldwide: its impact on operational efficiency, passenger experience and security, as well as the challenges that emerged with its implementation.

So which airports are leading the charge, who is making costly mistakes, and who is getting things just plain wrong?

The Good: better efficiency and customer experience

Kicking things off, let’s look at London Heathrow and Amsterdam Schiphol, two of Europe’s busiest airports.

AI is aiding air traffic management at both, particularly during challenging weather conditions like fog or heavy cloud cover.

Usually these conditions would cause significant delays and pose risks, but AI-driven systems now allow these airports to monitor and manage air traffic with absolute precision. This ensures safe and efficient operations, even in low visibility.

This advancement not only improves operational efficiency, but also sets a new benchmark for safety in airport management.

Singapore Changi Airport has also embraced AI to enhance runway management.

Its AI system continuously scans the runway for hazards such as debris or wildlife, a task that was once manual and prone to error.

Using high-resolution cameras and machine learning, the system detects even the smallest anomalies, reducing the risk of accidents and disruptions.

Changi’s proactive use of AI in this area is setting a new standard, likely to be adopted by airports globally.

In the realm of customer service, AI is also helping airlines and airports improve passenger interactions.

United Airlines, for example, uses AI to send personalised text messages with real-time updates on potential weather disruptions, including live radar links. This approach reduces uncertainty and has led to higher passenger satisfaction.

Similarly, Lyon Airport has integrated AI into its customer service with a chatbot that assists passengers with common inquiries, such as flight status and terminal directions.

While the chatbot still faces challenges with more complex questions, its implementation highlights the growing demand for automated services, especially during off-peak hours when human staff availability may be limited.

Lyon Airport’s use of AI in this capacity reflects a broader trend toward automation, meeting the evolving expectations of modern travellers.

The Bad: still not getting things right

Although AI has proved to be beneficial, there are areas where AI implementation has faced challenges, highlighting the need for ongoing refinement and adaptation.

Air Canada deployed an AI chatbot to handle inquiries about bereavement policies.

Unfortunately, the chatbot’s inability to grasp the sensitive nature of these inquiries led to frustration and misinformation regarding refunds for tickets.

This case highlighted that AI, while useful in many contexts, is not yet capable of addressing the emotional and nuanced needs of passengers dealing with sensitive issues, such as the death of a loved one.

In such scenarios, human interaction remains irreplaceable. AI should be used to complement, not replace, human empathy and understanding.

The Ugly: ethical concerns

However, the issues can run deeper still. One AI implementation that has raised ethical concerns is that of facial recognition.

San Francisco International Airport faced backlash over this, due to fears that AI’s inherent biases could lead to those within ethnic minority communities facing discrimination.

There were also data and privacy concerns that culminated in a citywide ban on the technology by public bodies.

Although the airport was exempt from the ban and still utilises facial recognition technology today, the controversy highlighted the unease around utilising AI within airport security.

To successfully integrate AI in a way that won’t see widespread backlash, airports must learn from past mistakes and ensure that human interaction is maintained where necessary – AI is not ‘The Great Replacer’ as it just isn’t built to understand the subtleties of human nuance.

The flight-path forward

We are at a point now where AI is out of its teething stages and there are actual use cases within airports and on airlines.

However, thoughtful implementation, continuous improvement, and a balance between technology and the human touch are essential to getting the very best out of it.

For example, before implementing AI technology, the US’s TSA (Transportation Security Administration) faced several challenges, including inefficiencies in security screening processes, long wait times, and human error in identifying threats.

The manual nature of these procedures led to inconsistent results, with security staff sometimes missing prohibited items or misidentifying benign objects as threats.

Additionally, the reliance on human judgement alone often caused bottlenecks, contributing to passenger frustration and delays.

The adoption of AI aimed to address these issues by enhancing threat detection accuracy, speeding up the screening process, and reducing false positives.

AI served to enhance these capabilities – and didn’t outmode anyone within airport security in order to do so.

It comes down to the fact that AI is a tool to enhance human functions, not a replacement.

Successful integration involves keeping humans in control, using AI to support and streamline processes while ensuring there’s always a human in the loop to manage exceptions and handle complex situations.

By learning from both successes and failures, airports and airlines can navigate the complexities of AI deployment and create a more efficient, customer-friendly experience for years to come.

That way, the sky truly is the limit.