1. Was the iBorderCtrl system used to perform border checks on real travellers?

iBorderCtrl was a research project, researching and developing new technologies. As the system it developed is only a research prototype, it cannot be used for actual border checks. Testing was carried out in an isolated manner (akin to a laboratory-test process), meaning that the iBorderCtrl system was not linked with any law enforcement databases or live systems.

Following a phase where the system was tested to assess its usability with isolated tests, the system needed to be tested further to validate whether the developed technologies were capable in principle of functioning properly under realistic conditions and real-life scenarios. To achieve this, test pilots were performed; some of these took place at selected border crossing points, with only members of the iBorderCtrl Consortium organisations, which have been invited as research participants, so as to avoid any real-life effects on persons outside the iBorderCtrl Consortium. As regards the research participants from the Consortium, their participation would not and did not result to any negative consequences for them in real-life, as it aimed at scientific research and their personal data were treated with integrity and care according to H2020 ethics requirements and GDPR standards.

2. Was the system used at European borders in Greece, Latvia and Hungary (or any other Member States)?

The test pilots were conducted at two selected border crossing points – one each in Hungary and Latvia, as well as in KEMEA and TRAINOSE premises in Greece. As outlined above, the reason for parts of the pilots taking place at border crossing points was to simulate real conditions in various scenarios, while the test pilots were fully isolated from actual border checks.

3. How were fundamental rights protected during the testing of the system?

The Consortium, from the beginning of the project, recognized the sensitivity of the proposed approach and technologies from an ethical point of view and took appropriate measures throughout the lifespan of the project, by continuously monitoring and making all the necessary efforts in order to ensure the respect of the fundamental rights of each person directly or indirectly involved. As outlined above, the test pilots were isolated from other proceedings at the border crossing points, whereas they were conducted exclusively using personnel from the iBorderCtrl Consortium; issues with respect to discrimination, human dignity, etc. therefore could not occur for ordinary members of the public.

Equally, the protection of the rights and freedoms of the research participants recruited from the Consortium, was guaranteed by the iBorderCtrl ethics framework and personal data protection policy. Thus, as required by the human rights framework and the H2020 ethics standards for the protection of the right to privacy and personal data, and for the participation in research, the participants were asked to provide informed consent prior to participating in the pilot tests and to the processing of their personal data. In this regard, they were informed about the research objectives and the data processing, their rights as data subjects, as well as their right to withdraw from the pilot tests at any time and ask for their data to be deleted. 

The invitation and selection of participants was designed avoiding discrimination against potential participants, e.g. ensuring a proper gender balance.

Personal data collected in the pilot tests was not shared with any third parties (such as law enforcement agencies) and was securely deleted after the project testing phase concluded in August 2019.

Last but not least, the iBorderCtrl system was seeking to avoid discrimination, by design, whenever possible. This being said, the system included an avatar in both genders and with different languages relevant to the use-cases fostered in the project.

4. How mature is the iBorderCtrl system following its development by the project?

iBorderCtrl was a ‘proof of concept’ research project as opposed to being concerned with mature product development. Thus, its core focus was on researching and promoting technology at a conceptual level, while seeking to evaluate its potential benefit for society and its impact on privacy and fundamental rights of individuals. During the project runtime, a first prototype of the iBorderCtrl toolkit was developed. However, due to its nature as a prototype and the technological infrastructure on EU level, it is not currently suitable for deployment at the border. Further development and an integration in the existing EU systems would be required for a use by border authorities.

While most components of the iBorderCtrl toolkit can be considered to be fully mature, it remains the case that novel technologies such as AI require extensive research on technological, legal and societal level before these are utilized in real-life applications. This also entails careful assessment and appropriate mitigation of potential adverse effects on the persons affected by their use. While the iBorderCtrl Consortium was able in principle to show the functionality of such technology in border checks, it is also clear that ethical, legal and societal constraints need to be addressed prior to any actual deployment. This is, however, a matter beyond the scope of iBorderCtrl. Indeed, some matters of concern (such as potential algorithmic bias and lack of transparency of decision-making) are also not specific to the field of border checks, but a common area of concern in AI research.

5. Will the system be used at the European border in the future?

As of now, iBorderCtrl is only a research project funded by the EU under the H2020 programme. How far the system, or parts of it, will be used at the border in the future will need to be defined. It should be also noted that some technologies are not covered by the existing legal framework, meaning that they could not be implemented without a democratic political decision establishing a legal basis. At the time of doing so, appropriate safeguards would also need to be considered, e.g. as proposed by the iBorderCtrl Consortium, to ensure the system operates with full respect for protect human rights.

6. What are the benefits and risks stemming from Artificial Intelligence?

The capabilities of Artificial Intelligence (AI) and the huge potentials it offers are subject to intensive research, not only in iBorderCtrl, but across the whole academic world and various disciplines. The iBorderCtrl Consortium believes that AI will greatly shape the future, not only in border checks, but in general. Consequently, it is important to accumulate knowledge on the advantages and disadvantages of the technology, and have an open debate on the issues as the basis for an informed democratic decision of society at large. This will help to ensure that such a system will only be used at the border if it provides fairer and better results than the current system, solely relying on human beings. For instance, an AI-based system with high accuracy might prove to decrease the risk of discrimination and other infringements of fundamental rights if designed and implemented properly.

With regard to iBorderCtrl, it can be concluded that the automatic deception detection system (ADDS), which relies on AI, poses various risks with regard to fundamental rights. As the iBorderCtrl cannot provide 100% accuracy, there is always a risk of false positives (people being falsely identified as deceptive) and false negatives (criminals being falsely identified as truthful).  This is true of any decision-making system, including those where classifications are made by humans. This might also lead to a stigmatisation or prejudice against affected persons, for instance when talking to a real border guard afterwards.

To this extent, it is vital to implement all measures and safeguards to combat  discrimination against travellers through the system. This includes the functioning of the algorithms as such (which must not utilize discriminatory factors and must correct and balance any discriminatory outcomes), as well as the data which has been used to train and test the system. In a real-world environment, strong safeguards would need to be applied, such as supervision through an ethics board for the algorithm, the training and test data and a constant monitoring of the system’s results. Also, border guards using such a system would need to be trained and sensitized accordingly.

Another challenge is to ensure a proper protection of human dignity. As the ADDS system utilizes human-machine interaction, it is important to make sure that the system treats people with respect to their dignity. People unfamiliar with such technology must not face any disadvantages, and should have the opportunity to rely on established procedures with human beings. In the light of this, it is suggested that technological solutions such as iBorderCtrl are configured to provide just an additional offer for travelers to use if desired, helping to reduce waiting times (similar to the eGates at airports).Therefore, it is important to notice that AI can assist but not replace human conduct, so the final decision is to be made by the Border Guards or other competent authority.

A final point to emphasise is that, prior to actual implementation of such a system, a concrete legal basis would be required,  which under the current applicable European legal framework, this is arguable. In this regard iBorderCtrl, as a research project, aimed at elaborating new technologies and testing whether in principle they have the capabilities to enhance the quality of border checks for both travellers and border guards, rather than implementing new policies.

The above-mentioned findings of the iBorderCtrl project show that further research will be required to fully exploit the potential of AI technologies, and prior to there being any actual deployment of ADDS for real-world border control. At the same time, the project can serve as a basis, providing input for further research activities not only in the field of border checks, but Artificial Intelligence in general.