CEPOL calls for biometric interoperability, training for cross-border investigations
The European Union Agency for Law Enforcement Training (CEPOL) has published a practical research paper describing a new Person-Centric open-source intelligence (OSINT) approach relying on Artificial Intelligence (AI) and interoperability to solve the challenges that emerge during investigations.
Titled “Artificial Intelligence and Interoperability for Solving Challenges of OSINT and cross-border Investigations,” the paper was written by former Egyptian Police Colonel Engineer Amr el Rahwan and tackles challenges like identity frauds, exchanging cross-border information, and the complexity of OSINT investigations.
The paper also mentions the Russian invasion of Ukraine and how it created the essential need for exchanging cross-border information to prevent, detect, and investigate terrorism and serious crime across Europe and neighboring countries.
As far as biometrics are concerned, the paper identifies various challenges connected with OSINT investigations.
“Although the EU-LISA will implement the interoperability framework in 2023, new challenges will emerge, such as investigating multiple-identity and identity frauds due to the different formats and structures of data,” reads the document.
For instance, fingerprints of a third-country national could be enrolled in one system with specific identity information. In contrast, the same fingerprints might be registered in another system using different identity information.
Face images also present challenges, according to the paper. Different photos of the same suspect could be enrolled in other systems and, because of the low-quality images or algorithms, could be categorized as two separate individuals. More training on facial recognition, therefore, is needed.
According to el Rahwan, these OSINT investigations are challenging because only officers with strong IT skills and backgrounds can obtain solid results from them. The newly introduced Person-Centric OSINT approach aims to solve these challenges by providing detectives and investigators with the necessary IT skills to operate AI systems within a more structured interoperability framework.
This training should include Named Entity Recognition (NER) technology to help officers understand the information that is useful for investigations while decreasing the focus on the less valuable or irrelevant information, as well as how to operate Natural Language Processing (NLP) systems to link different name variations of similar identities.
In related news, forensic examiners in the Netherlands recently discussed a procedure to acknowledge and try to mitigate subjectivity in face biometric applications.
