How can ChatGPT and similar tools revolutionize human-machine interaction at airports?

Considering the two key aspects in terms of travel retail, namely air traffic forecast and passenger profile, and taking into account that people are spending more time on personal devices, using the internet and social media, let’s take a closer look at generative AI and tools like ChatGPT. Do they have the potential to seize the opportunity and revolutionize not just travel retail but the entire commercial offerings at airports? 

While use cases are currently still limited, the number is expected to increase exponentially in the coming years, with customer experience being the primary focus. We have identified potential advantages and disadvantages of this exciting new technology.

• Reducing stress and anxiety of passengers: The potential to predict and identify disruptions during travel and intervene with real-time monitoring capabilities can drastically reduce passengers’ anxiety and stress, especially in the event of disruptions that lead to longer waits for security screening and regulatory checks. Generative AI and tools such as ChatGPT could be used to create an itinerary analyzer that incorporates multiple real-time information sources and helps optimize the passenger journey by providing alerts for traffic and schedule incidents.
  
  
• Multilingual support: Travel retail is all about serving to travelers from different countries and cultures, which often involves dealing with language barriers. Chatbots can be programmed to provide multilingual support and communicate with passengers in their preferred language. Its efficient ability to engage in interactive conversations enables it to adapt to tone and style, depending on the context, and provide relevant answers to passengers’ queries. This human-like interaction can help improve the passenger experience at airports. Google’s Bard, for example, supports the most commonly spoken languages, including English, Japanese, Korean, German, Arabic, Chinese, Hindi, and Spanish.
  
  
• Recommending promotions and offers at airports: Considering passengers’ previous shopping preferences, budget, purchase patterns, travel history, and specific requirements, chatbots supported by generative AI can act as a personalized marketing tool by sending targeted airport promotions and customized offers in real time. Such exclusive discounts, loyalty rewards, and incentives can encourage passengers to spend and serve as a personal shopping assistant.
  
  
• Pre-ordering and customization: Chatbots can facilitate pre-ordering services by allowing passengers to browse and compare available products, place orders in advance, and request product delivery either directly to the boarding gate, upon arrival, or to their home. In addition, they could make individual requests, such as an engraving on perfume bottles or personalized messages on gift items. This feature saves time at the airport and ensures that passengers have their desired products ready for pickup.
  
  
• Feedback and reviews: Airports can collect feedback from passengers on their shopping experience and product satisfaction. Online reviews or ratings can influence the purchasing decision of future travelers and contribute to building a positive reputation for the airport’s retail offerings. The software company Air AI, for example, is currently working on creating an AI assistant that has the potential to be a major player in the field of AI conversations. It is said to be able to have 10- to 40-minute-long phone calls that sound ‘like a real human’, with infinite memory, perfect recall, and the ability to autonomously take actions across over 5.000 applications[5].
  

• Upselling and cross-selling: With the support of AI technology, it is possible to suggest complimentary or enhanced products based on the passenger's initial purchase or browsing history. Highlighting related items or packages can encourage passengers to explore further options and make additional purchases, thereby increasing sales.
  
  
• Data collection and analysis: Generative AI can collect valuable real-time data on travelers' preferences, shopping behavior and feedback, providing airports and retailers with in-depth insights to make well-considered, data-driven decisions. As a result, airports can analyze this data to understand customer trends, optimize product offerings and improve the overall commercial strategy in collaboration with retailers. It can also efficiently support with content creation for airport commercial marketing strategies by generating images, slogans, campaigns, and more.
  
  
• Scalability: By being able to respond to multiple passenger interactions simultaneously, chatbots are highly scalable. They can easily manage large volumes of passenger enquiries without the need for additional resources, which can be very beneficial for airports, especially during peak travel seasons.
  
  
• Continuous improvement: As chatbots continue to collect more and more data, they could continuously learn and improve their responses and recommendations – the same goes for airports. Through generative AI, airports can gain accurate and valuable insights on, e.g., changing market trends, customer preferences, and industry updates. Having this detailed information allows airports to target their customers’ needs and ensure they deliver the best service possible.
  
  
• Maximize utilization of commercial spaces: Generative AI tools can help optimize passenger flows in terminals and analyze their movements using eye-tracking software solutions to study their visual attention, perception, and behavior. This not only allows optimal use of commercial spaces, but also the creation of virtual commercial spaces with diverse environments.

• Adaptability and real-time data updates: Generative AI models require regular real-time data updates and re-training to stay relevant and up-to-date. In the airport and travel retail industry, where conditions and offerings can change rapidly, relying solely on pre-trained generative AI models may limit the ability to adapt to real-time updates, resulting in outdated recommendations or information. ChatGPT’s data, e.g., does not extend beyond September 2021.
  
  
• Limited creativity and innovation: While generative AI can generate content based on existing patterns and examples, it might struggle with true creativity and innovation. In the travel retail sector, where novelty and fresh ideas are essential, relying purely on AI-generated suggestions or recommendations could lead to a lack of truly unique and groundbreaking offerings.
  
  
• Lack of human touch: Generative AI can produce realistic content, but it may lack the genuine human touch and authenticity that customers often seek. People often value personal experiences and connections that allow them to fully enjoy the essence of a destination or the unique offering of a retail store. Generative AI technology reduces opportunities for human interaction, which can be a disadvantage for customers who appreciate the expertise, empathy, and personal attention provided by human employees.

• Inability to handle complex situations: Although chatbots like ChatGPT can handle many routine customer requests, it may have difficulties with more complex or nuanced situations. Its responses are based on patterns learned from its training data and may not have the critical thinking or problem-solving skills of a human agent.
  
  
• Language and cultural limitations: Chatbots may not fully understand the nuances of different languages or cultures, resulting in possible miscommunication or misunderstanding with customers from diverse backgrounds. It might also have trouble with slang, colloquialisms, or regional variations, which can impact the accuracy and effectiveness of responses.
  
  
• Lack of product familiarity: In travel retail, customers often have specific questions or preferences regarding products like duty-free items, travel accessories, or destination-specific offers. Chatbots training data might not include comprehensive information on specific products, leading to limited or generic responses.
  
  
• Security and privacy concerns: When using chatbots, customer interactions and data may be processed and stored, potentially raising privacy and security concerns. It is critical that appropriate data protection measures are in place to safeguard sensitive customer information. If AI-generated content is primarily based on biased or incomplete data, it may unintentionally reinforce stereotypes or discrimination, causing unfair practices or exclusion of certain customer groups.

Building customer trust in AI-generated content will take time and requires transparent communication about the limitations and benefits of this new technology. Nevertheless, we cannot ignore that adoption of artificial intelligence is more than a passing fad; it is an important investment in the future of the airport and travel retail segment. It will be interesting to see how airports respond to generative AI – will they develop their own apps, chatbots and services like a digital shopping assistant, or will they rely on information from third-party generative AI solutions such as ChatGPT or Bard.

Regardless of which approach an airport chooses, it must ensure a seamless interface between the airport’s physical commercial offerings and the digital platform to enhance the overall passenger experience. The challenge will be to strike a right balance between using artificial intelligence tools such as ChatGPT and maintaining human interaction to continue to deliver cutting-edge services. At Munich Airport, we embrace the advantages of automation while ensuring that our customers can reach a human being whenever they may need or want to.