A Bottom-up Look at Mutual Trust and the Legal Practice of the Aranyosi Test

: This contribution offers an insight into the legal practice of the Aranyosi test during the EAW proceedings in seven Member States, an outcome of the research conducted during the ImprovEAW Project. Only the executing judicial authorities of some Member States do trigger the test. Member States are roughly differentiated between those having facilities with usually bad or usually good detention conditions, promoting antagonistic relationships instead of equal partnership. The lack of streamlining of the communication when supplementary information is requested, the lack of common standards and approach towards guarantees lead to further misunderstandings and frustration. The findings of this research have revealed the importance of departing from a pure legal understanding of mutual trust and follow a more empirical, experiential or bottom-up concept. Mutual trust is not only a legal concept, but it underpins the legal culture of the cooperation and collegial attitudes of authorities towards one another. This expression of mutual trust remains quite undiscovered: how is miscom-munication affecting mutual trust? Do judicial authorities of legal systems express collegiality to one another? How do cultural aspects and preconceived ideas regarding the quality


Introduction
Seven years have passed since the judgement by the Court of Justice of the European Union (CJEU) on Aranyosi and Căldăraru, in which, for the first time, it introduced exceptions to the unyielding concept of mutual recognition in European Arrest Warrant (EAW) cases. 1 With this judgement and those ensuing, the CJEU established a test where the executing judicial authority may postpone and eventually block surrender if there are serious suspicions that the issuing Member State (MS) will not respect the rights stemming from Article 4 of the Charter of Fundamental Rights of the European Union (Charter) and Article 3 of the European Convention on Human Rights (ECHR). 2 One particular aspect of the test is that the executing judicial authorities should engage into a dialogue with the issuing judicial authorities and request further information.
Yet only recently has scholarship focused on the actual national practice of the MS in applying the Aranyosi test. 3For instance, it is unclear Review of European and Comparative Law | 2023 Vol.54, No. 3 whether executing judicial authorities are obliged to trigger it and whether they actually do.It is not self-evident why judicial authorities would suddenly and gladly accept obstacles to a legal procedure that has run more or less smoothly for years if it were not for Aranyosi.Legal practice might differ regarding the actual tactics in requesting supplementary information, the content, the style of the responses and their evaluation by executing judicial authorities.
More generally, the executing judicial authority casts doubt upon the legal system of another MS.That arguably creates an uneven dynamic between the issuing and executing authorities: the executing judicial authority is empowered to review the standards in the executing MS; the issuing judicial authority on the other hand is interested in being taken seriously as an equal, trusted partner, but also aims at having the EAW executed swiftly without any delay.This dynamic might lead to a climate of mistrust and disappointment that goes both ways.The adverse impact of the Aranyosi test on mutual trust has been frequently addressed in prior scholarship and has been the source of much criticism. 4It has been particularly troubling to reconcile it with the previously supported strong concept of mutual trust.
In this contribution, my aim is twofold.Firstly, I present key observations from the practice in selected MS of how the Aranyosi test is used.My research is not exhaustive but it is based largely on the results of the EU-funded Project ImprovEAW, concluded in July 2022, where, inter alia, the application of the Aranyosi test had been assessed in seven MS, namely Belgium, Greece, Hungary, Ireland, the Netherlands, Poland and Romania. 5The research in those systems originates from the work output of the Project's partners arising from case files, interviews with judicial authorities and personal expertise.
Publishing, 2019), 437-454; Adriano Martufi and Daila Gigengack, "Exploring Mutual Trust through the Lens of an Executing Judicial Authority.The Practice of the Court of Amsterdam in EAW Proceedings, " New Journal of European Criminal Law 11, no. 3 (2020):  282-298.4   Weyembergh and Pinelli, "Detention Conditions, " 35-37.5   Renata Barbosa et al., Improving the European Arrest Warrant (Den Haag: Eleven, 2023); Barbosa et al., European Arrest Warrant: Practice in Greece, the Netherlands and Poland  (Den Haag: Eleven, 2022 Secondly, I shall use the lessons from the practice as a steppingstone to rethink the concept of mutual trust within the context of the Aranyosi test and the newly established dynamic in the relationship of authorities.One important takeaway is that mutual trust is a true sentiment and expectation amongst national judicial authorities.A vote of non-confidence could be potentially felt negatively.I will argue that one must approach mutual trust also as empirical and experiential concept (not only legal) and design measures that are bottom-up, targeting the judicial culture and improving the sentiments of collegiality and communication between authorities.Next to this, I will explore the idea to interpret the Aranyosi test as a risk management tool.In that respect, executing and issuing authorities are tasked with assessing and redressing possible risks generated by mutual trust.Inter alia, this approach could alleviate -to some degree at leastsensations of a broken relationship or inferiority on the part of the issuing MS and could promote a more focused and proactive attitude towards the challenges that individuals face.

The Design of the Aranyosi Test
The Aranyosi test is not meant as an automatic, casual part of surrender proceedings. 6Once the national court decides to commence it, there are two steps to be followed.First, the court must investigate whether detainees in the issuing MS run a real in-abstracto risk of being subjected to inhuman or degrading detention conditions.Such a general risk would exist in the case of "deficiencies, which may be systemic or generalised, or which may affect certain groups of people, or which may affect certain places of detention, " 7 not incidental or occasional violations.To assess such an abstract risk, the court may rely only on objective, reliable, specific and properly updated information.
If a general risk is established, the second step focuses on the ad hoc case and the existence of an in concreto risk for the specific requested person.To that end, the executing judicial authority engages in a dialogue with the issuing judicial authority. 8Supplementary information (Article 15(2) FD EAW) 9 must be requested from the issuing MS to inquire into the conditions of the facilities where the specific individual will be held.Both requesting for information and responding to such a request are obligations of the competent authorities.Once a response is received, the executing judicial authority must rely on the information acquired from the issuing MS but it may also rely on any other information available. 10Should that assessment result in finding a real in concreto risk for the requested person, the executing judicial authority must postpone the execution of the EAW until it receives information to mitigate the established risk within reasonable time, and if not, the procedure should be brought to an end.
Two observations can be made at this stage.First, the CJEU leaves large discretion to the executing authority in triggering and evaluating the potential risks. 11To be fair, the CJEU has contextualised the use of the test in subsequent jurisprudence and recently the EU has provided more guidelines on the standards of detention conditions, being points that I will refer to later.But the function of the test relies heavily on diverse factors: the national EAW procedure (e.g.how elaborate the procedure is, how much space for presenting evidence there is); the quality and expertise of the defence attorney required in convincing the court to trigger the test; whether national judges are well aware of the CJEU jurisprudence; how a national court perceives its role within the EAW procedure and the attention it puts on efficiency and mutual trust or, conversely, fundamental rights. 12econd, the procedure, especially the second part of the test, incorporates a dialogical aspect that is embedded in mutual trust.An obligation to engage in a dialogue echoes the most basic expectation in any relationship between equal partners.Accordingly, the procedure and content of 8 Martufi and Gigengack, "Exploring, " 283.   9   Council Framework Decision 2002/584/JHA on the European arrest warrant and the surrender procedures between Member States (OJ L 190, 18.7.2002).the information exchanged but also the overall quality of such communication are all elements demonstrating the effectiveness of the Aranyosi test as well as the impact that such interaction could have on the day-to-day practice of the EAW.

Practical Application: Experience from Seven MS
In the EU-funded Project ImprovEAW, the application of Aranyosi was investigated by looking at the practice and empirical evidence stemming from case files and interviews with practitioners.The aim of the project was broader, namely to improve the use of the EAW form, but the research was partly focused on the application of Aranyosi.In the following pages, I will share a few prominent observations.

The Aranyosi Test Is Triggered Mainly by Courts of Specific MS
Perhaps the most direct observation from this Project is the inconsistency in the application of this test.There is a variable geometry: some MS perform the test frequently when executing EAWs (the Netherlands, Ireland and Belgium), and some MS do not (Greece, Poland, Hungary and Romania). 13ndeed, from 2016 to September 2019, Dutch courts established an in abstracto real risk in 94 cases concerning Bulgaria, France, Hungary, Portugal, Romania and the UK.In 56 out of those cases, the surrender procedure continued since no in concreto risk had been established.In 38 out of those cases, a concrete risk was established and in 8 cases the person was surrendered after sufficient guarantees had been given. 14The Netherlands applied the Aranyosi with respect to: Belgium, Bulgaria, France, Greece, Hungary, Italy, Lithuania, Sweden, Poland, Portugal, Romania and the United Kingdom. 15he Irish executing authority had been busy with detention conditions and applying a test similar to the Aranyosi test already before Aranyosi. 16rish courts provided a series of cases in a similar vein to Aranyosi 13 Barbosa et al., "Improving, " 208-211.14 Barbosa et al., "European, " 192-197.15 Barbosa et al., "European, " 192-193.16   Minister for Justice, Equality and Law Reform v. McGuigan [2013] IEHC 216.Comparative Law | 2023 Vol. 54, No. 3 developing even further the application of this test. 17The detention conditions of several legal systems have troubled the Irish courts, e.g.Romania, Lithuania, the UK, Poland. 18urprisingly, no cases were reported for this Project when Greece, Poland, Hungary and Romania acted as executing judicial authorities, even when dealing with the EAWs stemming from the MS with known challenging prison conditions. 19Whereas it cannot be said that the judicial authorities of those MS have never triggered the test, this result is puzzling.

Reasons for Inconsistent Application of Aranyosi
Why such an uneven trigger of the Aranyosi test?For once, the Aranyosi-jurisprudence, being relatively new, might not have dwelled in all systems.It could be that executing judicial authorities might not be familiar with the CJEU jurisprudence in this regard, especially in those systems where more national courts can act as executing judicial authorities. 20That could be even more true for defence attorneys: the defence carries the load to at least prompt the court to launch the test.From a European Parliament study, the question has emerged whether the executing authorities should perform the first part of the test ex officio. 21Similarly, the Dutch courts had considered briefly the possibility of introducing a step 0, namely to investigate preliminary evidence that is insufficient to trigger the test but somewhat intriguing; eventually it has been abandoned as idea, because insufficient evidence means just that and nothing more. 22Nevertheless, the role of executing authorities within the national concept of criminal procedure can be tricky given the burden left to the defence, a concern raised by the Irish and Dutch authorities: sometimes the Dutch courts will help the defence by gathering information ex officio. 23nother explanation could stem from the purpose of the Aranyosi test all together: MS with known challenging detention conditions (such as Greece, Romania and Hungary) cannot commonsensically complain about detention conditions elsewhere. 24It sounds illogical and hypocritical to complain when you have your finger in the pie. 25Such explanation could result into a limited application of the Aranyosi test: is it meant de facto to be triggered only by those MS with usually good detention conditions?
The detention conditions at the executing MS are irrelevant for triggering the test, but in practice this hesitation could be real.
There is the possibility that, in some MS, the attitude of courts might be leaning more towards the side of an unconditional mutual trust or simply prioritise the effectiveness of the EAW over fundamental rights in some cases, especially if seen from the point of managerialism and efficiency.EAW procedures have been running for years relatively successfully.Why change a winning horse?Naturally, the EAW proceedings at the national level have gained their own pace and automatism.Some courts might be reluctant since the Aranyosi test is a taxing judicial exercise, workload is high and there are tough time-limits.
All of those are educated assumptions and more research is needed into the actual practice.Still, those findings are awkward.Aranyosi relates to an absolute fundamental right.There is something inherently wrong with the fact that being arrested in some MS might lead to higher protection of this absolute right than being arrested in others.
Importantly, the Aranyosi test is a purely judicial exercise without a proper legal basis: the FD EAW does not include a ground for refusal 22 Barbosa et al., "European, " 198.23 Barbosa et al., "European, " 199.24   Weyembergh and Pinelli, "Detention Conditions, " 35.concerning fundamental rights.This is reported to already cause confusion to national courts as to how to phrase their refusals to execute. 26The amendment of the FD EAW to include such formal refusal ground has been put forward and debated. 27Yet, the whole point with introducing mutual recognition is to move away from the older extradition model.Already it is argued that those recent developments in the CJEU jurisprudence create a regime far more stringent than the previous extradition framework. 28

The Aranyosi Test Can Damage the Trust between the Judicial Authorities
This uneven legal practice gives rise to a division between some MS that usually invoke Aranyosi (the Netherlands and Ireland) and MS usually at the receivers' end of requests because of poor detention conditions (Greece, Poland, Hungary and Romania). 29This polarity is not absolute as some MS belong to both categories (e.g.Belgium). 30But grouping MS into the good and bad students of the class is the opposite of what you want in the Area of Freedom Security and Justice where mutual trust reigns.
More specifically, such picture was observed when looking at how those judicial authorities receiving plenty requests outlined their experience.Several of those prosecutors and judges in the aforementioned MS reported feeling mistrusted and they assessed some requests as exaggerated, pedantic (e.g.whether the defendant will be allowed to smoke tobacco) or as an expression of superiority. 31 impartiality of courts (and not detention conditions but they particularly exemplify the situation), Poland refused to execute EAWs from the Netherlands arguing a lack of impartiality of the Dutch authorities, meaning as impartiality towards Polish cases. 32On the other hand, the executing authorities are also brought into the difficult position to judge the facilities and standards of other legal systems. 33

Problematic Dialogue
The way in which the MS request and provide supplementary informationbeing unstandardised and left for the MS to figure out -does not help improve communication.For example, the Greek issuing authorities reported that questions often arrived one-by-one prolonging the procedure. 34Frustration with supplementary information is also witnessed outside the field of Aranyosi as other supplementary information requested is often irrelevant or exists already in the EAW form. 35hether or not excessive questions are asked depends on the standards of detention conditions, a topic without current EU harmonisation.Misunderstandings may occur as to which questions are (ir)relevant.The CJEU has made a consistent reference to the case law of the ECtHR: all physical aspects are relevant (e.g.personal space, sanitary conditions, freedom to move, 3m 2 minimum with certain exemptions). 36A few other established guidelines are as follows: the executing judicial authority may not use higher national standards as a benchmark; 37 the executing judicial authority must not request supplementary information on all prisons of the issuing MS, only on the actual and precise facilities where the requested person will likely be detained, including on a temporary basis; 38 to establish a real 32 Barbosa et al., "European, " 299.33   Weyembergh and Pinelli, "Detention Conditions, " 31.34 Barbosa et al., "European, " 77.35 Barbosa et al., "Improving, " 199-201.risk in concreto, the review must be comprehensive and not limited to only manifest inadequacies, meaning that supplementary information should allow gathering information for such comprehensive review; questions regarding a legal remedy are not necessarily irrelevant but a legal remedy in the issuing MS to challenge detention conditions does not suffice to exclude a real risk of violation. 39ven if useful, those guidelines are insufficient to direct judicial discretion and ensure proper communication, which is also criticised in scholarship. 40When supplementary information is requested, several issuing authorities choose to offer a type of guarantee that the requested person will be kept in appropriate facilities.Offering such a guarantee expedites surrender.But with no clear and common detention standards it becomes unpredictable for the issuing authority which guarantee would be satisfactory.The following case perfectly exemplifies this problem: two concurrently pending Greek EAWs were put under consideration before two different German courts and exactly the same supplementary information was provided by the Greek authorities regarding the same prison (exactly the same numbers, capacity and conditions), yet one German court found the prison satisfactory but the other did not. 41n a similar vein, sometimes even after offering a guarantee the request was refused, leaving the issuing authorities puzzled. 42Such a situation could either result from an inadequate guarantee given by the issuing judicial authority or because the executing judicial authority did not rely on it.Several factors could influence the reliability of the guarantee: too abstract or laconic or subject to conditional terms or provided by a non-judicial authority which casts doubt regarding its reliability, given the recent case law on this matter. 43Additionally, there are various types of guarantees given, apparently: depending on the national circumstances some issuing authorities might promise (not) to use a specific prison or give a general promise to use 41 Barbosa et al., "European, " 78-79.42 Barbosa et al., "European, " 78-79.43 Barbosa et al., "Improving, " 219.  a facility with adequate requirements. 44Without any common standards, its reliability depends fully on the executing authority, causing uncertainty.
Such unpredictability could foster a culture of gambling on the part of the issuing authority.From the Greek practice, for example, two attitudes emerge when answering supplementary information requests, depending on whether it is possible to comply with the requests and how important the ad hoc case is.If so, then the attitude is to take full responsibility to comply with fundamental rights and take all necessary measures to ensure appropriate facilities, picking prison facilities that are "Aranyosi-proof ".If, on the other hand, there is simply no space in such facilities that are deemed undoubtedly appropriate, and/or if the case is not too pressing, the Greek issuing authorities will simply communicate the conditions of the prison allocated for this case.And then the burden to decide whether those detention conditions are (in)sufficient rests with the executing authority.That latter approach has been referred to by some Greek practitioner as a "gamble" that sometimes "works", 45 meaning that borderline cases could sometimes pass the test because the executing judicial authorities might give way to the pressure of mutual trust, the need to bring cases to justice and execute requests.One has to wonder: can the EAW procedure regain its credibility and reliability as mechanism of cooperation?

A Bottom-up Approach to Mutual Trust
All in all, one observes that communication and perception of the relationship between authorities play a significant role for mutual trust.Bilateral communication between authorities could be less or more prolific depending on the MS involved.The research conducted within the framework of the ImprovEAW Project, while not exhaustive as far as the application of the Aranyosi test is concerned, indicates that mutual trust is not only a normative legal concept.It is also something that judicial authorities experience, an experiential concept that informs their decisions in more subtle ways and underpins the attitude that authorities have when cooperating: asking too many irrelevant questions, casting doubts on the quality of legal The predominant view towards mutual trust in criminal matters approaches it as a legal and normative concept.It is a postulate governing the Area of Freedom Security and Justice enabling authorities to cooperate despite their differences.Mutual trust has been used as a tool for the purpose of achieving pluralism of legal systems and forming the basis on which mutual recognition of judicial decisions can exist. 46oncomitantly, mutual trust has had an ambivalent, paradoxical nature: it is an existing postulate based on common values but also an objective to be reached in itself, and an obligation arising for achieving another purpose, i.e. mutual recognition.For example, in the Preamble of the FD EAW mutual trust is presented as a pre-existing postulate based on common values: "The mechanism of the European arrest warrant is based on a high level of confidence between Member States." 47 But in the Hague Programme, mutual trust is featured as a goal in itself, something to be created by means of legal measures. 48For AG Bot in Kossowski, though, mutual trust is none of the above but a consequence of mutual recognition: The intention of the EU legislature, in adopting the principle of mutual recognition, was to overcome the almost insurmountable difficulties which had been encountered, due in particular to the failure of efforts to approximate national laws in advance.The Court followed the legislature by giving due effect to the principle in its case-law.The phrase used must therefore be understood as meaning that mutual trust is not a prerequisite for the operation of mutual recognition, but a consequence which is imposed on MS by the application of that principle.In other words, the application of the principle of mutual  (OJ L 190, 18.7.2002),recital 10.48   The Hague Programme: strengthening freedom, security and justice in the European Union (OJ C 53, 3.3.2005).
Review of European and Comparative Law | 2023 Vol.54, No. 3 recognition requires the MS to place trust in each other regardless of the differences in their respective national laws. 49l three narratives have coexisted all along creating a confusing picture but also showing that mutual trust has heterogenous nature and cannot be exhaustively explained by means of a single narrative. 50s a presumption it has been almost irrebuttable for years. 51Until Aranyosi, the CJEU had been reluctant to allow discretional powers to executing authorities in assessing the compatibility of the EAW with fundamental rights; 52 although some discretion was afforded in the areas of law lacking EU harmonisation. 53This uncritical view of the presumption of mutual trust (which has received criticism over the years) has persevered. 54his has been the case even if in modern instruments (post-EAW), such as the European Investigation Order, grounds for refusal on fundamental rights were in fact introduced. 55To be fair, such strict application of mutual trust was accompanied by a series of Directives on procedural rights.Yet there is no harmonisation on the measures of detention per se, such as the requirements for arrest and pre-trial detention. 56nyhow, trust has been superimposed as a given.In that sense, it has also been a static concept, it exists no matter what the reality is.However, national constitutional courts on many occasions sang a different tune, opposing to such static understanding of trust; in some of the MS even grounds for refusal related to fundamental rights were introduced in national law, contrary to EU law. 57ith Aranyosi (and the ensuing case law), the CJEU relativised the presumption of mutual trust and introduced this test as a form of control, a decentralised fundamental rights mechanism.This signalises a paradigm shift from an automated, presumed, uncritical trust to a dialogical process between authorities, where trust is checked and earned when doubts exist.From a static conception of "trust is there" we have moved to a dynamic conception of "trust is being built in a continuous feedback loop".
Looking at mutual trust as a dynamic concept has two advantages: it allows for the open exploration of what constitutes mutual trust and what influences it.Maybe Aranyosi has harmed mutual trust, but perhaps in the long run, such dialectic process will strengthen mutual trust: systems will understand one another better or the peer-review and pressure mechanism will perhaps encourage MS to raise the standards of human rights protection or reduce the use of EAWs.Mitsilegas, for example, examines the versatile and complicated nature of mutual trust by exploring its relation to fundamental rights: mutual trust has strengthened fundamental rights but it can also be strengthened by the harmonisation of fundamental rights or it can also be an obstacle for the protection of fundamental rights. 58he second advantage is that, in ad hoc EAW proceedings, attention is paid to the legal practice.Whether or not there is mutual trust depends on the MS in question and their communication and cooperation.Mutual trust could be strengthened if this communication is fluent, with clear mutual expectations of engagement.Trust is also built in a bottom-up way.
More bottom-up empirical research is required to investigate the use of Aranyosi.After all, mutual trust is a concept known to socio-political science, anthropology, economy, and psychology as an important trait that underpins human relations. 59It transcends law.It would be very interesting to gain overview into the actual dynamics of relationships of judicial authorities: their perception of collegiality, the judicial culture of their cooperation and how to foster it and their culture for professional work ethics.Additionally, while we know something regarding the responses of the issuing authorities, we know very little of what executing judicial authorities do with the received information, how they make decisions and which factors they put attention to.For example, the Belgian executing authorities sometimes request information regarding the ways the executing MS has responded to ECtHR judgements, implying that this is a factor to consider in their decision.Rogan's research reveals a possibly pivotal role for national and supranational bodies tasked with the monitoring and supervision of prisons, as the existence of such bodies might play an invisible but important role in the decision-making process. 60o be clear, my claim is not that top-down measures have no place.The lack of common standards in detention conditions has been a problem-maker. 61Establishing EU common standards of detention conditions would improve mutual trust.In December 2022, the European Commission published a Recommendation on the rights of suspects in pre-trial detention including material detention conditions. 62It proposes standards for pre-trial detention but also material conditions with extensive reference to several aspects of the life in prison including vulnerable detainees such as women and those with vulnerable health conditions.The Recommendation is certainty a large step towards the right direction.
But would harmonisation be enough to improve mutual trust and the proper applicability of Aranyosi?One would require more bottom-up measures that address the judicial culture of the "coal face" practice.61 Barbosa et al., "Improving, " 315.62   Commission Recommendation on the procedural rights of suspects and accused persons subject to pre-trial detention and on material detention conditions.Brussels, 8.12.2022C(2022) 8987 final.
to the legal order of the other.When there is a presumption of compatibility not corresponding to the reality, there are potential risks generated: violations of individual rights, inconsistencies/inefficiency of instruments, tensions with legal principles of national orders, and other disadvantages either affecting the state or individuals.Eliminating those risks becomes a necessary part of a dynamic concept of mutual trust.And the Aranyosi test does exactly that. 67ne advantage with that view is its accurate depiction of the test.Rizcallah explains how such a test already fits into the overall outlook of generic risk assessment tools. 68The Aranyosi test can then become more methodical and transparent in its function.Take for instance the first step of a generic risk analysis which is to identify the risk and whether it is worthy of action.When identifying risks in the field of mutual recognition, it would be important to first illustrate whether the risk is pertinent to an important foundational value (or human right) and whether such risk is actually serious. 69In fact, Aranyosi has been expanded to aspects of fair trial and the lingering question is whether its scope could be expanded to other fundamental rights as well. 70In EDL, the CJEU held that when the requested person suffered from a chronic and serious illness, the executing authority could request supplementary information and assurances that the detention conditions would accommodate the health condition of the requested person. 71n the pending case of GN, the AG has opined that the executing authority could refuse the surrender of a mother of minor children when the executing authority is not absolutely certain, after requesting supplementary information and assurances, that the issuing state will respect the rights of the child during detention. 72Those are all generated risks created by mutual trust.But should those risks be mitigated?
A second advantage arising from Aranyosi as a risk management exercise is that it changes the narrative of the relationship of the authorities.Instead of monitoring and supervising the detention conditions in the issuing 67 Rizcallah, "Le principe, " 539.state, both authorities address and minimise some risks created during the EAW proceedings.When Aranyosi is triggered, the modality is not one of control and supervision but of division of labour in addressing a risky situation for individuals.Both authorities in their role should accordingly make all efforts possible to mitigate that risk.This approach could help those MS with already problematic detention conditions to trigger the test themselves when executing EAWs.This is echoed also in AG Bot's Opinion: "(…) We must also not forget that the issue here is to prevent a risk, not to find and penalise an infringement (...)." 73 More generally the risk management approach promotes a more proactive policy for the potential risks of mutual recognition and the position of individuals in general.Perhaps a part of the process in producing legislation in the field of mutual recognition should be to conduct a risk assessment of the potential risks that mutual trust would create for the individuals subjected to new instruments and how to best offset them.One example is the post-effect of the EAW: it is unclear whether those surrendered with an EAW tend to have worse treatment as far as pre-trial detention is concerned in the issuing MS, or they stand a similar chance to be granted bail.Additionally, it is not clear whether the Aranyosi test itself creates two-class citisens within the same MS: those surrendered under the EAW and after assurances, who are consequently kept in better facilities, and the rest of the detainees. 74

Conclusion
In this contribution, I have presented key observations regarding the practice of Aranyosi arising from the ImprovEAW Project.There is inconsistency in triggering the test and the division between MS, which promotes antagonistic relationships instead of equal partnership.The lack of streamlining of the communication when supplementary information is requested, the lack of common standards and approach towards guarantees lead to frustration.The findings of this research have revealed the importance of abandoning a purely legal conceptual framework of mutual trust and focusing on a more empirical, experiential or bottom-up concept.Mutual trust creates a legal culture of judicial cooperation during EAW proceedings, which remains quite undiscovered: how is miscommunication affecting mutual trust?Do judicial authorities of different legal systems express collegiality to one another?How do cultural aspects and preconceived ideas regarding the quality of legal systems influence mutual trust and the decisions of national authorities?If mutual trust is not always a given anymore and courts are allowed to debate it, those are good questions to be addressed.There should be more research conducted in order to reveal the perspective of judges and prosecutors who operate the EAW proceedings.Accordingly, some suggestions have been made to improve the cooperation and the establishment of rapport when supplementary information is requested.Finally, I have supported a more neutral view towards the Aranyosi test.As opposed to looking at it as a supervisory mechanism, I have explored the idea to approach it as a risk management tool.In my view, such approach helps both sides to take responsibility to avert ad hoc risks, instead of experiencing Aranyosi as a one-sided testing moment.Such approach puts the real problem in the centre: the risks created by mutual trust for individuals.
25 TP Marguery, "Towards the End of Mutual Trust?Prison Conditions in the Context of the European Arrest Warrant and the Transfer of Prisoners Framework Decisions, " Maastricht Journal of European and Comparative Law 25, no.6 (2018): 712-713.

59Eveline
Brouwer and Damian Gerard, "Mapping Mutual Trust -an Introduction, " in Mapping Mutual Trust: Understanding and Framing the Role of Mutual Trust in EU Law, eds.Evelien Brouwer and Damian Gerard (EUI Working Paper: 2016), 1. 60 Mary Rogan, "What Constitutes Evidence of Poor Prison Conditions after Aranyosi and Căldăraru?Examining the Role of Inspection and Monitoring Bodies in European Arrest Warrant Decision-Making, " New Journal of European Criminal Law 10, no. 3 (2019): 209-226.

46
Valsamis Mitsilegas, "Conceptualising Mutual Trust in European Criminal Law: the Evolving Relationship between Legal Pluralism and Rights-Based Justice in the European Union, " in Mapping Mutual Trust: Understanding and Framing the Role of Mutual Trust in EU Law, eds.Evelien Brouwer and Damian Gerard (EUI Working Paper: 2016), 25.