Artificial intelligence and sustainable tourism: major challenges and opportunities

Artificial Intelligence (AI) has the potential to transform the tourism sector and move it towards a more sustainable model, but it also raises major energy issues. At the 2nd Luxembourg Tourism Summit 2024, an event focused on AI and sustainable tourism, enriching discussions provided a better understanding of how AI could be responsibly integrated into tourism. This post aims to provide an initial assessment of the relationship between AI and sustainable tourism.

1- AI energy consumption: A crucial challenge for sustainable tourism

Figures on data center energy consumption and the carbon footprint of AI models show a strong upward trend. In 2024, all data centers consumed around 1,000 terawatt-hours (TWh) per year, i.e. almost 2% of global energy consumption, or the equivalent of the total consumption of France and Germany combined.

The production of CO₂ to drive a large AI model remains high, but new data also clarifies the impact depending on the methods and infrastructure used. Giants like Microsoft, Amazon and Google are investing heavily in renewable energy, which could reduce this footprint in the long term, but critics point out that transparency efforts and monitoring standards will be needed to avoid “greenwashing”.
These data show the need to integrate IA’s responsible energy strategies for sustainable tourism, and to favor more energy-efficient models and renewable energy sources to limit the footprint of technologies.

At this level, four (4) main avenues emerge:

1) Efficient use of generative AI: A single query on ChatGPT, for example, would consume around 2.9 Wh and emit 1.38 g of carbon, i.e. around ten times more energy than a typical Google search. To minimize emissions, users should focus generative AI on targeted, essential uses, avoiding superfluous processes. Every AI query consumes energy, and it is possible to reduce this consumption by adjusting resources precisely to the needs of each task.

2) Energy optimization: It is essential to develop AI models that consume less energy. Advanced graphics processing units (GPUs) and optimized algorithms are promising avenues.

3) Use of renewable sources: Data centers should turn to solar, wind, hydroelectric (possible in Quebec) or other renewable sources to reduce their footprint.

4) Balancing costs and benefits: as discussed at the Luxembourg summit, it is essential to strike a balance between the sustainability gains brought about by AI and its energy cost.

2. Sustainable resource management with AI

AI offers solutions for distributing visitor flows more evenly and better managing tourist sites. One example is the “First Index of Overtourism” set up by Evaneos and Roland Berger, which objectively measures the phenomenon of overtourism in 70 of the world’s top 100 tourist destinations. The system uses real-time data to assess the impact of tourism on various territories. By identifying the destinations most affected by overcrowding, AI helps to regulate the influx of tourists. For example, the platform has already planned to stop selling summer trips to Mykonos and Santorini from 2025, due to the very high environmental and social impact during this season.

By using AI to analyze criteria such as tourist density per inhabitant or per square kilometer, seasonality, and infrastructure sustainability, managers can anticipate and better plan for peak periods. This helps avoid overcrowding at popular sites, while promoting less crowded periods or alternative destinations.

AI thus contributes to sustainability by enabling intelligent resource management and providing businesses and local authorities with tools to better organize visitor reception, while preserving the environment and reducing the negative impact of mass tourism.

Overtourism is not just a question of the number of visitors at a given time, but also of the destination’s capacity to manage this mass of visitors, which is not necessarily included in the index.

In Quebec, this approach could be beneficial in relieving pressure on certain popular natural sites, while encouraging tourists to explore other, less-visited regions, thereby reducing environmental impact and preserving local ecosystems.

3. Personalizing the tourism experience and sustainable marketing

AI makes it possible to personalize visitors’ experiences according to their preferences, an asset for encouraging responsible behavior without compromising their satisfaction. For example, a tourist interested in history might receive recommendations for cultural visits outside the traditional circuits, while a nature lover might be directed towards environmentally-friendly activities, such as visits to less-frequented national parks.

AI can also contribute to “sustainable marketing” by highlighting environmentally-friendly options in personalized recommendations, such as eco-friendly accommodation, restaurants using organic produce, or low-carbon transport routes. More responsible choices thus become an integral part of the offer, encouraging visitors to adopt more sustainable behavior.

4. Real-time environmental monitoring

AI systems, coupled with intelligent sensors, enable real-time monitoring of the environmental impact of tourism. For example, cities like Barcelona use sensors to monitor air quality, noise levels and population density in tourist areas. Similarly, Yellowstone Park uses AI-based systems to monitor wildlife movements and prevent ecosystem disturbances.
This technology provides managers with valuable data to adjust their practices and minimize their ecological footprint. It allows them to better organize the reception of visitors, while preserving the environment and limiting the negative effects of mass tourism on natural sites.

5. Using predictive AI for sustainable development strategies

Predictive AI is a strategic tool for identifying sustainability trends in the tourism sector. By analyzing data from multiple sources, such as social networks, online reviews and booking data, AI can detect tourist behaviors and preferences, as well as emerging sustainability trends.

For example, the Auvergne-Rhône-Alpes region in France uses AI-analyzed sustainable tourism indicators to track trends and adapt its offerings to the expectations of environmentally conscious travelers. In Quebec, destination managers could adopt similar practices to adapt their tourism offerings in line with these new trends, ensuring more responsible development and meeting the expectations of today’s travelers.

6. Challenges and collaboration for sustainable tourism

Discussions at the Luxembourg Tourism Summit highlighted four key challenges for optimizing AI in the sustainable tourism sector:

1. Resource optimization: Algorithms can optimize visitor flows and distribute resources more evenly, reducing pressure on popular destinations.
2. Stakeholder engagement: Collaboration between governments, companies, NGOs and local communities is essential to co-create sustainable solutions that benefit all stakeholders.
3. Innovation and research: Investing in emerging technologies such as predictive AI is crucial to proactively managing the challenges of sustainable tourism.
4. Awareness-raising and education: Educating tourism professionals about the benefits and challenges of AI is essential for responsible use and to encourage sustainable travel practices.

Conclusion: AI, a lever for sustainable tourism

Energy-intensive AI presents major challenges, but also undeniable advantages in terms of sustainable tourism. By optimizing resources, facilitating personalized, eco-responsible experiences and enabling better management of environmental impacts, it is positioning itself as a powerful tool for a greener future. However, it is essential to reduce our energy footprint and adopt balanced strategies to minimize our negative impacts. Working together with the various players in the industry, we can make AI a real catalyst for sustainability in tourism.

Sources:
https://www.mckinsey.com/industries/private-capital/our-insights/how-data-centers-and-the-energy-sector-can-sate-ais-hunger-for-power

https://www.datacenterdynamics.com/en/news/global-data-center-electricity-use-to-double-by-2026-report/