Security expert, Paul Trotter, weighs in on the pros and cons of AI in travel safety decisions.
With artificial intelligence (AI) capabilities evolving at a rapid rate, the opportunities for businesses to enhance their systems and processes are continually growing. In the realm of travel risk assessment, this potential is especially promising, offering greater efficiency and insight. However, when it comes to protecting travellers and ensuring their safety, how much should we rely on AI compared to the expertise of human professionals?
Understanding the opportunities AI presents and the associated risks is crucial in striking a balance between technological efficiency and human judgment. To get a better understanding of how AI is used in the current travel risk management landscape we spoke to Paul Trotter, Security and Intelligence Lead at World Travel Protection, who provided valuable insights and real-life examples.
How is AI currently being used in travel risk assessment?
Technology can positively contribute to travel risk assessment by being used as a tool for data analysis. It can rapidly process and clean large volumes of information, making it easier for security experts to identify patterns and make quick assessments about events and their potential impact on travellers.
“We do this by utilising systems that seamlessly identify events based on a targeted keyword search collected across tens of thousands of sources of information. Not only does this enhance our efficiency and speed of detection, but it also allows us to make recommendations faster to those in need,” explains Trotter.
However, risk identification cannot rely on technology alone. It’s important for experts to assess all the incoming information provided by tools such as AI and verify it for reliability and accuracy.
“This analysed data can then be further refined by our team, who overlay predicted risks and outcomes to create a comprehensive risk assessment. By combining expert-verified travel risk alerts with our Travel Assist app and its geolocation functionality, we can automate the delivery of these alerts to travellers within an affected area. This ensures that we provide only relevant information to travellers, empowering them to make informed decisions when needed.”
What are the risks of AI misinterpreting or overlooking critical risk factors?
1. Ability to interpret data
In its current state, AI is simplistic in its approach to delivering risk assessments. While it can identify risks to a degree, based on patterns and trends, it cannot make judgement on the risks present and provide tailored mitigation strategies.
AI lacks the ability to account for many factors that come with human knowledge, such as adding context, judgement, and critical thinking – qualities that are vital in the field of travel risk assessment.
For example, consider a political unrest alert in a destination such as Thailand. While AI can flag that a demonstration has turned violent, it does not consider the secondary impacts or broader implications. Nor can AI accurately identify the causal factors that cause the protest to turn violent, limiting the efficacy of such a system to be predictive. It is up to human experts to interpret the severity of the threat, understand local context and assess whether this will have a wider impact on travellers and what can be done to mitigate risks.
2. Understanding the individual travel risk profile
When it comes to assessing risk, it’s also crucial to understand each traveller’s individual risk profile and identify the risks associated with them. This is another area that AI cannot be relied upon, as it only looks at the facts provided and does not factor in individual needs.
For instance, AI may be able to analyse traveller information and acknowledge that someone has asthma, but it does not consider the availability of salbutamol or other asthma medications at their intended destination.
“Oversights like this may seem minor on paper, but they can be catastrophic in practice. This is why World Travel Protection’s team of experts consults with clients regularly during the development of travel risk management services to ensure we meet the specific and tailored needs of travellers,” shares Trotter.
In the above context, it’s up to experts to understand each traveller’s nuanced needs, such as the amount of salbutamol required daily, and align this with local legislation on imports, the availability of medication in country and the reliability of potentially required medical support.
3. Filtering out misinformation
One of the biggest risks of using AI is its inability to distinguish facts from misinformation. Because AI systems rely on large datasets to generate insights, if the data used is outdated, biased, or inaccurate, the resulting analysis can be false, misleading and skewed.
Similarly, AI has a tendency to return false information of its own creation due to programming quirks. From miscalculating numbers or misidentifying details, to returning answers that are not grounded in fact, users are continuously at risk of false or misleading information, intentional or otherwise. This highlights the ongoing need for careful verification when using AI as a source of information.
The other consideration is the impact of geopolitics and the host-nation of where the platform is developed or based, or from where the information is sourced. Different views on the existence (or not) of a country, the risk level in a destination, and the types of risks travellers may be exposed to can all change based on the politicisation of information which is difficult to counter when we are unaware of the originating source.
Striking a balance between AI-driven insights and human expertise
Despite common belief, AI technology is still very much in its infancy and lacks the emotional and intuitive intelligence that is so necessary when it comes to comprehensive risk assessment.
“Human involvement is still very much critical. Whilst we can employ various AI and automation technologies to enhance our efficiency, our risk analysis and assessments factor in a wide range of variables across human, environment and traveller task terrains” says Trotter.
AI merely provides one piece of the puzzle, which is completed by the knowledge and insights of human experts.
“School excursions are a great example of why we operate this way. Larger groups require much more detailed risk assessments due in large part to the increased number of variables that need to be considered. When you include children into the mix, any risk assessment and advice must provide information that meets all of their individual risk needs whilst also providing realistic and achievable mitigation strategies that teacher-supervisors can enact,” shares Trotter.
While AI can easily analyse large quantities of information, it cannot apply its findings in context. Furthermore, it does not have the intuition required to determine which mitigation strategies will work and which can be effectively applied. It’s important to consider how to strategically use AI while leveraging human understanding of group dynamics, experience, and cultural intricacies.
Learn how we can support your business
Curious to learn more how we can support your business travellers with our in-house team of medical and security experts? Fill out the form below to schedule a chat with our team today.
