Features

AI in Border Management: Implications and Future Challenges

By Dr. Erdal Düzdaban

The rapid advancement of information technologies has prompted significant discourse regarding the potential societal benefits of artificial intelligence (AI). By definition, AI refers to the capability of devices, guided by computer systems, to perform tasks traditionally associated with intelligent beings. Research on artificial intelligence dates back to the 1940s. A key focal point in this field has been the development of systems that can generalize, classify, and learn from past experiences, ultimately applying this knowledge in ways that are relevant to human needs.

Significant advancements are being made in the development of artificial intelligence (AI) technology, not only to enhance human economic welfare but also to improve applications in security domains. In the realm of crime prevention, data collected in compliance with legal standards are analyzed within the constraints set by law. This legal framework guides the interpretation of data and informs the implementation of preventive measures aimed at reducing criminal activity. Law enforcement agencies leverage existing and interconnected data through the application of artificial intelligence (AI) to conduct comprehensive data analysis. This process allows them to identify meaningful relationships between various datasets and visualize these connections, thereby enhancing their ability to combat crime effectively. In the fight against crime, analytical approaches vary by country, influenced by the available data capacity.

The data utilized encompasses open sources as well as legally obtained information pertinent to criminal investigations. However, if legally collected data fails to translate into actionable insights for operations, it hinders the establishment of a uniform application standard across countries regarding data security and retention periods. Generally, law enforcement agencies are more effective in combating crime in border regions due to their significantly greater capacity and resources compared to organizations tasked with border protection. Often, border protection agencies share responsibilities with customs organizations operating in the same area.

Although legal frameworks outline the specific geographic responsibilities and operational protocols for these entities, issues related to authority, accountability, and coordination frequently arise, impeding effective collaboration. Border agencies routinely conduct risk assessments to prevent cross-border crimes and address potential threats in border regions. Within the Schengen framework, there is a degree of data sharing among member countries, and some nations participate in global and regional intelligence exchanges.

However, despite established internal guidelines regarding the reliability, preservation, usage, and coordination of shared data, achieving a cohesive approach remains problematic. Questions often arise about who can access this data, the extent of its utilization, and how it can be effectively integrated into operational strategies.

For instance, in drug enforcement initiatives, the outcomes of cooperative efforts, such as controlled delivery operations involving multiple countries, frequently have limited relevance beyond the participating nations. Uncertainties persist regarding the sharing of operational data with other countries, the sanctions imposed on individuals involved in these operations, and whether post-incarceration processes are communicated to relevant authorities across jurisdictions.
Additionally, information concerning individuals on watch lists in counter-terrorism efforts—despite not being subject to active criminal investigations—can still be utilized as operational data, raising further questions about privacy and the scope of its application. In essence, it is crucial to define the types of data subject to artificial intelligence (AI) and establish robust control mechanisms for this data.

At the European Union level, the General Data Protection Regulation (GDPR), effective from March 25, 2018, provides a comprehensive framework for data privacy across all member states. This regulation establishes rules governing the use of personal data and its transfer to other countries and international organizations. In addition to general data processing guidelines, it also addresses specific scenarios, such as legal obligations in the public interest and legitimate interests, ensuring a balanced approach to data protection.

In border areas, the question of whether to utilize intelligence information in addition to data on individuals previously involved in operations is a distinct issue that warrants examination. In order units primarily require information for tactical and operational assessments. However, engaging in long-term, probabilistic strategic evaluations, along with optimizing capacity for these assessments, may constrain the possibilities and capabilities of the relevant units.

Accordingly, the application of AI technology in data analysis for border management and security primarily supports tactical and operational decision-making rather than facilitating long-term projections. Its effectiveness is largely contingent upon the specific data it utilizes, enhancing immediate responses in the context of ongoing operations.

Let’s consider an example of AI applications in data analysis within a hypothetical country (X). Border units in country (X) can access various data sources, including information from other security agencies, financial transactions through banks, watch lists, data from regional unions, and the INTERPOL database. Additionally, base station information can be incorporated into this data.

Currently, AI is utilized to query and list information based on specific criteria. While this analysis can provide descriptive insights into past events, it has limitations. For instance, if a person is detected communicating from the same base station as an individual on a watch list—whose connection to terrorism lacks supporting evidence—this alone would not suffice for making informed decisions. Relying solely on such data could infringe upon the individual’s right to free movement.

AI technology, as it stands, lacks the capacity for human-like analysis of the aforementioned data. Given the current pace of technological development, it appears unlikely that AI will achieve a level of understanding comparable to human reasoning in the near future. Although AI can establish connections based on available data, it cannot replicate the nuanced decision-making that humans can provide. In the context of border management and security, if the analytical capabilities of AI are confined to existing data, concerns regarding privacy and personal data protection will inevitably arise.

In conclusion, while technological advancements should enhance physical operations in border management and security, they should not be the sole determinants in decision-making processes.

Dr. Erdal Düzdaban held various ranks within the General Directorate of Security of the Republic of Türkiye, dedicating his career to combating organized crime and terrorism. He also served as a contracted expert in border management and security for international organizations, including the OSCE secretariat and its field offices in Kosovo, Tajikistan, and Turkmenistan. His experience extends to UN Peacekeeping Operations in Kosovo and the IOM Nigeria Office, where he made significant contributions to enhancing security and cooperation on a global scale.