International Protection Bill 2026: Report and Final Stages Seanad Éireann — 2026-03-25 ============================================================ Alice-Mary Higgins (IND), National University of Ireland The placement of these amendments is somewhat imperfect. This is another example of what happens when a guillotine is applied. We have amendments that try to address the same issue elsewhere in the legislation, but we believe they are of such crucial significance that it is important that they are addressed and discussed. I have said again and again that the fact of giving 1.2 minutes per amendment does not allow for best practice, and I believe it will lead to a number of flaws in the legislation. That is why we have had to seek to tackle the same issues in a couple of different ways. We should not have legislation that needs to be front-loaded because we assume that if we try to address a flaw at a later stage of the legislation, we will never get to discuss it. This amendment relates to the question of how facial image data is being used. My colleague will speak to this as well. When we spoke on Committee Stage about the reports indicating issues and concerns on the huge number of false identifiers that we have seen where facial recognition is being used, and the fact that it has been shown to be far less accurate with different ethnic groups and different skin colours, the Minister suggested that those were almost teething problems, that they are improving and that the reports were five or ten years old. I read into the record the fact that we had heard from the UK that the data regulator there had concerns in this regard as recently as 2024. In 2024, we were hearing of extremely serious concerns about the impact that using that kind of technology could have in terms of reinforcing biases within a system, with those biases having very serious consequences for different people. Even the idea of how saying the reports from five or ten years ago show problems but the situation has improved points to the woeful experiment that was covered in the newspapers seems to be lost, in that immigration advice might be given through some kind of a chatbot. Technology that has been shown to be deeply flawed and is producing new problems all the time should not be allowed to test and improve itself on vulnerable persons. They should not be guinea pigs for an area that has consistently been highlighted as needing far better regulation and tighter restrictions on its use. The reports from as recently as a few years ago and previously point to the fact that this should not be overused and overrelied on. We talked it through last time that if a person was wrongly identified, there did not seem to be a proper appeal mechanism for them to say they had been falsely identified by these algorithms. The Minister said that it would be relevant when a person's immigration application was considered. What access do people have to actually challenge the interpretation of that? It will simply be a matter of the facial recognition system suggesting who a person is while the person claims this is who they are. Where does the burden of proof sit? How will that be filtered out if we have a pattern of false identification issues? Earlier, regarding the age assessment of minors, the Minister made the case for why we needed age assessment. What we have spoken to is the idea that how age assessment is being done is a very significant concern. We point to the fact that medical age assessment or, indeed, age assessment processes are being triggered simply by officials who feel that they would like to have it triggered rather than it being an absolute last resort, which is what the UN Convention on the Rights of the Child has said. Our fear is that it is simply becoming standard. I am concerned that this Eurodac system and its facial recognition software might be used on minors. Will the Minister clarify that there are safeguards to ensure that is not the case? I also fear that it might be used by the authorities to suggest that what they consider a match for a person was found on the Eurodac system and they determine a person's age on the basis of that. I am very concerned about facial recognition being used in general, but particularly given its impacts in relation to certain ethnic groups - those with darker skin tones, who have been found to be disproportionately wrongly identified - and minors. I can come back to the Minister afterwards on these but they are core problems I see in what the Ombudsman for Children has identified as a new system, not the normal system where people were passing this first test, the one where people were effectively committing an offence or were being detained and forced to attend these screening centres. These screening centres may be relying on technology that reinforces biases and has very serious consequences, particularly for certain ethnic groups and minors. Will the Minister assure me that it will not be used as an argument in relation to the question of age assessment? --- Source: Houses of the Oireachtas. Licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). The Official Report is revised after first publication; the fetch timestamp below identifies the version quoted. Record URI: https://data.oireachtas.ie/akn/ie/debateRecord/seanad/2026-03-25/debate/main Retrieved: 2026-08-14T04:54:46+00:00 Sitting date: 2026-03-25