SafeTravellers promises to develop a “unique platform based on several ground-breaking technologies” to help Border Authorities and Law Enforcement Agencies identify and verify EU and Third Country National citizens through “multi-modal” biometric solutions. These are meant to combat identity and travel documents fraud at EU borders.
According to its Cordis page, the ambition is to enable an uninterrupted border crossing experience: “With SafeTravellers, border crossings become seamless, using a distributed European Multi-Biometric Data Space to facilitate secure identity checks without compromising personal data.”
While the current lack of project documents in the public sphere prevent us from understanding what this “unique platform” will actually consist of (What kind of “multi-modal” biometric solutions are to be developed? What is a “European Multi-Biometric Data Space”?), we do know, thanks to Biometric Update reporting, that the system will be based on “an integrated software platform that combines pre-registration, document verification and biometric authentication” to be developed by Idemia Germany. “This solution will be available as a smartphone app for travelers and border control agents, as well as through static versions for eGates and border kiosks,” the company said, according to the digital news outlet.
A (blurred) snapshot of SafeTravellers’s architecture shared on the project’s LinkedIn page seems to confirm that the smartphone constitutes the heart of the proposed system.
The only public deliverable we could consult (D7.1 Dissemination and Communication Plan) contains the “appropriate” messages to be disseminated to SafeTravellers’s target groups. Beside LEAs, Border Agencies and Coast Guards, these include international standardisation bodies, governmental organisations, the scientific community —plus, of course, “citizens and travellers”.
To them, the project should convey these simple messages:
– “Identity fraud is widespread and should be addressed by new technologies”;
– “SafeTravellers helps to reduce identity fraud and provides a new frictionless, efficient and secure experience without queues and limited security personnel interaction”;
– “The new biometrics-based identification is optional and purely based on consent”;
– “SafeTravellers protects the biometric data provided”;
– “SafeTravellers provides faster and more convenient border crossings for travellers.”
A set of technologies will be developed over the course of SafeTravellers. These include:
– “Threat Intelligence Based on Unusual Events/Patterns” (which, according to the redacted Grant Agreement we obtained from the Research Executive Agency, must “understand the person that is crossing the border, meaning if they are under duress”, possibly hinting at some form of emotion recognition)
– “Frictionless Identification for EU/TCN (third-country national, ndr) citizens”
– “Multi-Modal Biometrics Fusion”
– “XAI (explainable AI, ndr) based Recommendation Engine”
– “Novel Mechanisms for Preventing Look-Alike and Makeup Attacks”
– “Morphing Based Malware prevention”.
The project’s Cordis page further adds that artificial intelligence will be featured in:
– “AI-Driven Risk & Identity Assessment: Fuses multiple biometric modalities with AI to assess risk and improve security decision-making”
– “Identity Fraud Detection: Uses AI and machine learning to identify forged or altered identity documents”.
Just like the security industry and several other EU-funded projects, SafeTravellers material obsessively repeats the mantra of “seamless”, “frictionless” border crossings thanks to biometric technologies ad AI. In fact, the whole project is premised on the idea that, since “Identity and travel documents are increasingly being counterfeited or tampered with”, these solutions are needed to efficiently contrast the current “rise in transnational crime such as human trafficking or terrorism”.
According to a first SafeTravellers flyer, the project also “incorporates privacy preserving mechanisms like homomorphic encryption to safeguard citizens’ biometric data”.
More details can be gleaned from a Biometric Update article published in April 2025, which adds “a mobile wallet that stores digital ID documents such as mobile passports, allowing travellers to control data through self-sovereign identity (SSI) management and enabling a more frictionless travel experience”. It also claims that the project will combine face recognition and fingerprinting to strengthen accuracy, while “Suspicious patterns can be detected through threat intelligence and AI-driven analytics”.
All of these solutions promise “Interoperability with EU Systems”, ensuring “seamless integration with EES and ETIAS”.
The project’s Grant Agreement further specifies that “SafeTravellers will also connect with several legacy systems such as INTERPOL’s databases and reference libraries, which allows conducting first-line checks and second-line inspection at land, seaports and border crossings”.
Concerning AI, the document writes that “sophisticated image processing and deep learning algorithms combined with state-of-the art tamper detection and localisation schemes will be developed to automatically understand and generate a risk score regarding the legitimacy of the identity and travel document,” while “innovative image processing and robust AI-based techniques will allow the detection of authentic document of look-alikes and authentic travel documents under false pretences.”
Additionally, “In order to further enhance the situational awareness of the border authorities and LEAs, sophisticated AI-based threat intelligence mechanisms extracting insights from online sources and the Dark/Deep web will be employed”.
All of these technological solutions promise to reduce human error (“because unlike machines, security practitioners can get tired and lose their focus”), delays and frustration. In fact, “SafeTravellers’s “Mobile Passport” “will allow citizens to seamless cross the EU borders without the need of old-fashioned documentation”, reads the Grant Agreement.
In general, the idea is to automate the security checks as much as possible, “thus removing the burden of continuous person checks in border areas and optimizing the operational processes followed by LEAs and Border Guards” — or, as put immediately after, thus “minimising the required user intervention to the absolute necessary level”.
Deliverable D7.1 features (p. 16) a Table illustrating “Sister projects and collaborations”. Mentioned EU-funded projects include:
– TENSOR, whose mechanisms “for biometric extraction sharing and storage cross border environments will be examined providing in SafeTravellers a starting point for the creation of the Multi-Biometric European Dataspace”
– iFLOWS, to explore and further improve its “multi-tier toolkit enhancing intelligence extraction in border and customs areas”
– iMARS, to build on top of its “Morphing and manipulation attack detection solutions”
– ODYSSEUS, to leverage “The developed tools for secure identity verification allowing citizens to cross borders without any intervention by just leveraging their smartphones combined with strong and continuous identity verification”
– PERSONA and METICOS, to implement the “integrated impact assessment method” and “solutions for addressing the technology acceptance and efficiency” they developed
– CarMen, PopEye and EINSTEIN, in ways that are still to be determined.
Projects BODEGA and Smart-Trust are also mentioned as “related projects” in the Grant Agreement.
From an image shared on the project’s LinkedIn page we learn that four pilots are envisioned:
– Pilot 1 – Attacking biometrics at airports
– Pilot 2 – Fraudulent identity and travel document at sea and land borders
– Pilot 3 – New way of identification at land borders and within vehicles
– Pilot 4 – Enhanced identification on railway premises.
“Range of different scenarios is big, regulated by different laws, implemented on different technology stacks/capabilities”, reads the depicted slide, announcing that SafeTravellers will only “focus on some selected scenarios”.
According to the project’s redacted Grant Agreement we obtained from the Research Executive Agency following an access to documents request, “each project Pilot, prior to commencing the data processing activities, will carry out a Data Protection Impact Assessment”.
The same document specifies that the SafeTravellers platform aims at a technology readiness level (TRL) of 7 or higher, meaning that the system’s prototype must be demonstrated in an operational environment by the end of the project. This does not amount to immediately deployable solutions.
Nonetheless, the “Exploitation Plan” section for “Industrial Partners” states that they “will incorporate SafeTravellers technologies to their brochure of products ensuring the leading role in the European and Global market”. Also, it is explicitly stated that “The Border Authorities and Law Enforcement Agencies will adopt the developed technologies in order to improve the existing border control process and the safety of the border crossing points”.
The Grant Agreement also promises that “engagement of citizens and Civil Society Organisations will be established in order to also validate the project’s offered solutions”.
SafeTravellers’s idea for “seamless”, “frictionless” border crossings based on sophisticated biometrics, AI, and a digital pre-registration procedure is perfectly consistent with the logics and prescriptions of the EU ‘Migration Pact’, which will have to be implemented by mid-2026 by all Member States. Together with the technological outputs of current EU-funded projects, such as CarMen and PopEye, its solutions seem to be intended to realize the Pact’s dangerous vision for the future of EU borders.
Right from its About section, SafeTravellers claims to be placing “privacy and fairness at the center of its research”, considering “privacy, consent, and bias mitigation as core components of the development and exploitation of the proposed solution”.
And yet, the language adopted in the project description on Cordis repeatedly conflates migrants and criminals into the same disease, speaking of “an ‘epidemic’, fuelled by demand from asylum-seekers and terrorists using counterfeit documents”.
While foundational details of the project are currently lacking, we do know that several consortium members took part in previous EU-funded projects in border security. Among them, the Austrian Institute of Technology (which participates in BorderForce and PopEye), Idemia France (also member of the CarMen consortium), and the portuguese company Vision Box (also a member of ODYSSEUS and iMARS).
Only one deliverable, the project’s ‘Dissemination and Communication plan’, was publicly available some 18 months after the start of the project.
The Grant Agreement we obtained from the Research Executive Agency only partially sheds additional light on project details. In fact, several parts of it were either partially or completely redacted, including:
– the list of deliverables for Working Packages 3-5, together with their individual description; all deliverables description for WP8 are also fully redacted;
– target measurable key performance indicators for three project objectives;
– all description of the “Current State-of-the-Art” and progress beyond it;
– a large chunk of the “Methodology” section (pp. 14-21);
– a significant portion of the “Key target groups and audiences” section;
– the whole “Quality and Efficiency of the Implementation” section;
– the detailed Estimated budget.
Some 40 pages in a row (pp. 163-200 of the overall document we received) were entirely black.
These redactions are all the more noticeable given the explicitly pro-transparency stance taken by the project in its Grant Agreement: “The consortium embraces the vision that large and unrestricted access to knowledge is essential not only for the central role of knowledge and innovation in generating growth but also as a fundamental human value of scientific knowledge progress and dissemination”.
We did however understand that the project already underwent some significant changes, including its name (originally “SafeNomads”) and the composition of its consortium (two members were removed and replaced by other institutions).
Lastly, regulation and social acceptance are considered barriers to adoption. “The adoption of the platform and its underlying technologies is”, as for most other projects, “hindered by legislative barriers in national level” — including “(Biometric) data confidentiality, locality and protection posing potential restrictions in data usage and sharing across countries”. It is explicitly stated that “SafeTravellers will liaise from the beginning with policymakers, legislation agencies and standardisation bodies”. “Social resistance against information sharing between LEAs” is also considered a “social barrier”.
