1.5.2008   

EN

Official Journal of the European Union

C 110/1

1.

The draft Proposal for a Council Framework Decision on the use of Passenger Name Record (PNR) data for law enforcement purposes was sent by the Commission to the EDPS for consultation, in accordance with Article 28(2) of Regulation No 45/2001/EC (hereinafter ‘the proposal’).

2.

The proposal concerns the processing of PNR data within the EU and is closely related to other schemes of collection and use of passengers' data, in particular the EU-US agreement of July 2007. These schemes are of great interest to the EDPS, who already had the opportunity to send some preliminary comments on the Commission's questionnaire on the intended EU PNR system, sent in December 2006 to relevant stakeholders (3). The EDPS welcomes the consultation of the Commission. According to the EDPS, the present opinion should be mentioned in the preamble of the Council Decision.

3.

The proposal intends to harmonise Member States' provisions on obligations for air carriers operating flights to or from the territory of at least one Member State regarding the transmission of PNR data to the competent authorities for the purpose of preventing and fighting terrorist offences and organised crime.

4.

Arrangements for transmission of PNR data for comparable purposes have been concluded by the European Union with the USA, as well as with Canada. A first agreement concluded with the USA in May 2004 was replaced by a new agreement in July 2007 (4). A similar agreement was concluded with Canada in July 2005 (5). In addition, negotiations are due to start between the EU and Australia for an agreement on the exchange of PNR data, and South Korea is also requiring PNR data from flights to its territory, without any plan for negotiation at European level at this stage.

5.

Within the EU, the proposal comes as an addition to Council Directive 2004/82/EC (6) on the obligation of carriers to communicate passengers' data known as API data, in order to combat illegal immigration and improve border control. This directive should have been transposed in national law of Member States not later than 5 September 2006. Implementation is however not ensured yet in all Member States.

6.

Contrary to Advanced Passenger Information (API) data that are supposed to help identifying individuals, PNR data mentioned in the proposal would contribute to carrying out risk assessments of persons, obtaining intelligence and making associations between known and unknown people.

7.

The proposal includes the following main elements:

8.

The proposal on which the EDPS is consulted is a further step in a movement towards a routine collection of data of individuals who are in principle not suspected of any crime. As mentioned above, this evolution is taking place at international and European level.

9.

The EDPS notes that also the Article 29 Working Party and the Working Party on Police and Justice have presented a joint opinion on the proposal (8). The EDPS supports that opinion. The present opinion emphasises and develops a number of additional points.

10.

While the opinion of the EDPS will analyse all relevant aspects of the proposal, it will concentrate on four main issues.

11.

Other substantive points will be identified in a last part, including positive steps in terms of data protection but also additional sources of concern in the proposal.

12.

In order to analyse the legitimacy of the proposed measures in accordance with fundamental data protection principles, and notably Article 8 of the European Charter of Fundamental Rights and Articles 5 to 8 of Council of Europe Convention No 108 (9), it is necessary to identify clearly the purpose of the intended processing of personal data, to assess its necessity and its proportionality. It should be ensured that no other means is available, that would be less invasive, to reach the envisaged purpose.

13.

The wording of the proposal and its impact assessment indicate that the objective is not simply to identify known terrorists or known criminals involved in organised crime, by comparing their names with those included in lists managed by law enforcement authorities. The purpose is to gather intelligence with regard to terrorism or organised crime, and more precisely ‘to carry out risk assessment of persons, obtain intelligence and make association between known and unknown people’ (10). The purpose stated in Article 3(5) of the proposal is, in the same line and firstly, ‘to identify persons who are or may be involved in a terrorist or organised crime offence, as well as their associates.’

14.

This is the reason invoked to explain that API data are not sufficient to reach the alleged purpose. Indeed, as already mentioned, while API data are supposed to help identifying individuals, PNR data do not have an identification purpose, but the details of the PNR would contribute to carrying out risk assessments of the persons, obtaining intelligence and making associations between known and unknown people.

15.

The purpose of the measures envisaged does not only cover the catching of known persons but also the locating of persons that may fall within the criteria of the proposal.

In order to identify these persons, risk analysis and identification of patterns are at the core of the project. Recital 9 of the proposal states explicitly that data must be kept ‘for a sufficiently long period as to fulfil the purpose of developing risk indicators and establishing patterns of travel and behaviour’.

16.

The purpose is thus described in two layers: the first layer consists of the global objective to fight against terrorism and organised crime, while the second layer includes the means and measures inherent to the achievement of this objective. While the purpose of fighting terrorism and organised crime appears to be clear enough and legitimate, the means used to reach this purpose leave room for discussion.

17.

The proposal gives no indication on the way patterns will be established and risk assessment will be performed. The impact assessment gives the following precision as to the way PNR data will be used: to run the data of passengers ‘against a combination of characteristics and behavioural patterns, aimed at creating a risk-assessment. When a passenger fits within a certain risk-assessment, then he could be identified as a high-risk passenger’ (11).

18.

Suspected persons could be selected according to concrete elements of suspicion included in their PNR data (e.g. contact with a suspicious travel agency, reference of a stolen credit card), as well as on the basis of ‘patterns’ or an abstract profile. Different standard profiles could indeed be constituted on the basis of travel patterns, for ‘normal passengers’ or ‘suspicious passengers’. These profiles would enable investigating further those passengers who do not fall within the ‘normal passenger category’, all the more so if their profile is associated with other suspicious elements such as a stolen credit card.

19.

Although it cannot be assumed that passengers would be targeted according to their religion or other sensitive data, it appears nevertheless that they would be subject to investigation on the basis of a mix of in concreto and in abstracto information, including standard patterns and abstract profiles.

20.

One could discuss whether this type of investigation would qualify as profiling. Profiling would consist of a ‘computer method making use of data mining on a data warehouse, enabling or intended to enable the classification, with some probability — and thus with some margin of error — of an individual in a specific category in order to take individual decisions towards that person’ (12).

21.

The EDPS is aware that there are ongoing discussions on the definition of profiling. Whether or not it is officially recognised that the proposal aims at profiling passengers, the main point at stake is not about definitions. It is about the impact on individuals.

22.

The main concern of the EDPS relates to the fact that decisions on individuals will be taken on the basis of patterns and criteria established using the data of passengers in general. Thus decisions on one individual might be taken, using as a reference (at least partially), patterns derived from the data of other individuals. It is thus in relation to an abstract context that decisions will be taken, which can greatly affect data subjects. It is extremely difficult for individuals to defend themselves against such decisions.

23.

In addition, the risk assessment is to be performed in absence of uniform standards of identification of suspects. The EDPS seriously questions the legal certainty of the whole filtering process, considering that the criteria against which every passenger will be scanned are so poorly defined.

24.

The EDPS recalls the jurisprudence of the European Court of Human Rights, according to which domestic law must be sufficiently precise to indicate to citizens in what circumstances and on what terms the public authorities are empowered to file information on their private life and make use of it. The information ‘should be accessible to the person concerned and foreseeable as to its effects’. A rule is ‘foreseeable’‘if it is formulated with sufficient precision to enable any individual — if need be with appropriate advice — to regulate his conduct’ (13).

25.

To conclude, it is notably because of these types of risks, that the present proposal needs careful consideration. While the general purpose to fight against terrorism and organised crime is in itself clear and legitimate, the core of the processing to be put in place does not appear to be sufficiently circumscribed and justified. The EDPS therefore urges the EU-legislator to clearly address this issue, before adoption of the Framework Decision.

26.

The intrusive character of the measures is evident, as shown above. On the other hand, their utility is far from being demonstrated.

27.

The impact assessment on the proposal concentrates on the best way to establish an EU PNR, more than on the necessity of such PNR. Reference is made in the assessment (14) to PNR systems in place in other countries, namely the USA and the United Kingdom. One can however deplore the lack of precise facts and figures related to those systems. ‘Numerous arrests’ are reported with regard to ‘various crimes’ in the UK semaphore system, without precision as to the link with terrorism or organised crime. No details are given with regard to the US programme, except that ‘the EU has been able to assess the value of PNR data and to realise its potential for law enforcement purposes’.

28.

Not only is there a lack of precise information in the proposal on the concrete results of such PNR systems, but reports published by other agencies, e.g. the GAO in the United States, do not confirm at this stage the efficiency of the measures (15).

29.

The EDPS considers that techniques consisting of assessing the risk presented by individuals using data mining tools and behavioural patterns need to be further assessed, and their utility be clearly established in the framework of the fight against terrorism, before they are used on such a wide scale.

30.

In order to appreciate the balance between the intrusion in the privacy of the individual and the necessity of the measure (16), the following elements are taken into account:

31.

Compliance with the proportionality principle implies not only that the proposed measure is effective, but also that the purpose envisaged by the proposal can not be reached using the less privacy invasive tools. The effectiveness of the intended measures has not been demonstrated. The existence of alternatives must be carefully assessed before additional/new measures are put in place to process personal information. According to the EDPS, such comprehensive assessment has not taken place.

32.

The EDPS wishes to recall the other large scale systems monitoring the movements of individuals within or at the borders of the EU, whether in operation or about to be implemented, including in particular the Visa Information System (17) and the Schengen Information System (18). While these instruments do not have as a main goal the fight against terrorism or organised crime, they are or will be to some extent accessible to law enforcement authorities for the broader scope of the fight against crime (19).

33.

Another example concerns the availability of personal data included in national police data bases — especially with regard to biometric information — in the framework of the Prüm Treaty signed in May 2005, that is being extended to all Member states of the European Union (20).

34.

These different instruments all have in common that they enable a global monitoring of movements of individuals, even if from different perspectives. The way in which they can already contribute to the fight against specific forms of crimes, including terrorism, should be subject to in-depth and comprehensive analysis, before deciding to establish a new form of systematic scanning of all persons leaving or entering the EU by plane. The EDPS recommends that the Commission conducts such an analysis, as a necessary step in the legislative procedure.

35.

In the light of the foregoing, the EDPS concludes as follows on the legitimacy of the proposed measures. Building upon different data bases without a global view on the concrete results and shortcomings:

36.

The fight against terrorism can certainly be a legitimate ground to apply exceptions to the fundamental rights to privacy and data protection. However, to be valid, the necessity of the intrusion must be supported by clear and undeniable elements, and the proportionality of the processing must be demonstrated. This is all the more required in case of extensive intrusion in the privacy of individuals, as foreseen in the proposal.

37.

It can only be noted that such elements of justification are missing in the proposal and that the necessity and proportionality tests are not fulfilled.

38.

The EDPS insists on the essential character of the necessity and proportionality tests developed above. They represent a condicio sine qua non to the entry into force of the present proposal. Any further comment of the EDPS in the present opinion must be taken in the light of this preliminary condition.

39.

The analysis below will concentrate on three points:

40.

Article 11 of the proposal states that ‘Member States shall ensure that the Council Framework Decision on the Protection of Personal data Processed in the Framework of Police and Judicial Cooperation in Criminal Matters (…) is applicable to the processing of personal data under this Framework Decision.’

41.

However, in spite of this provision it is not clear to what extent the data protection framework decision — an instrument under the third pillar of the EU Treaty — will be applicable to data processed by airlines, collected by PIUs, and further used by other competent authorities.

42.

The first step in the processing of personal data foreseen by the proposal is processing by the airlines, which are obliged to make PNR data available — using in principle a push system — to national PIUs. It seems from the wording of the proposal and the impact assessment (22) that data could also be transmitted in bulk by airlines to intermediaries. Airlines are primarily active in a commercial environment, subject to national data protection legislation implementing Directive 95/46/EC (23). Questions on the applicable law will arise when data collected are used for law enforcement purposes (24).

43.

Data would then be filtered by an intermediary (to be formatted and to exclude PNR data not included in the list of data required by the proposal) or sent directly to PIUs. Intermediaries could also be actors from the private sector, as is the case for SITA, operating in that sense in the framework of the PNR Agreement with Canada.

44.

When it comes to PIUs, responsible for the risk assessment of the whole amount of data, it is not clear who will be responsible for the processing. Customs and border authorities might be involved, and not necessarily law enforcement authorities.

45.

The subsequent transmission of filtered data to ‘competent’ authorities would probably happen in a law enforcement context. The proposal states that ‘competent authorities shall only include authorities responsible for the prevention or combating of terrorist offences and organised crime’.

46.

While moving forward through the steps of the processing, the actors involved and the purpose followed have a closer link with police and judicial cooperation in criminal matters. The proposal does not explicitly mention, however, when precisely the data protection framework decision will apply. The wording would even lead to think that it applies to the whole processing, and even to the airlines (25). However, the framework decision on the protection of personal data includes in itself some limitations.

47.

In this context, the EDPS fundamentally questions the fact whether Title VI of the EU-Treaty can serve as a legal basis for legal obligations on a routine basis and for law enforcement purposes upon private sector actors. Additionally, the question is relevant whether Title VI of the EU-Treaty can serve as a legal basis for legal obligations on public authorities which are in principle outside the framework of law enforcement cooperation. These questions will be elaborated further on in this opinion.

48.

The text of the Proposal for a Council Framework Decision on the protection of personal data processed in the framework of police and judicial cooperation in criminal matters contains at least two limitations which are relevant in terms of scope.

49.

In the first place, the scope of the data protection framework decision is well defined in the framework decision itself: it applies ‘only to data gathered or processed by competent authorities for the purpose of the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties’ (26).

50.

In the second place, the data protection framework decision is not supposed to apply to data processed purely at domestic level, but is limited to data exchanged between Member States and further transfer to third countries (27).

51.

The data protection framework decision also includes some drawbacks compared to Directive 95/46/EC, in particular a wide exception to the purpose limitation principle. With regard to this purpose principle, the proposal clearly limits the purpose of the processing to the fight against terrorism and organised crime. However, the data protection framework decision allows processing for wider purposes. In such a case, the lex specialis (the proposal) should prevail over the lex generalis (the data protection framework decision) (28). This should be made explicit in the text of the proposal.

52.

For this reason, the EDPS recommends adding the following provision to the proposal: ‘Personal data transmitted by airlines according to this Framework Decision may not be processed for purposes other than the fight against terrorism and organised crime. The exceptions foreseen with regard to the purpose principle in the Council Framework Decision on the protection of personal data processed in the framework of police and judicial cooperation in criminal matters do not apply’.

53.

As a conclusion, the EDPS notes a serious lack of legal certainty with regard to the data protection regime applicable to the different actors involved in the project, and in particular to airlines and other first pillar actors: be it the rules of the proposal, the rules of the data protection framework decision or the national legislation implementing Directive 95/46/EC. The legislator should make clear at what moment of the processing precisely these different rules will apply.

54.

The EDPS fundamentally questions the fact that a third pillar instrument creates legal obligations on a routine basis and for law enforcement purposes upon private or public sector actors which are in principle outside the framework of law enforcement cooperation.

55.

A comparison could be made here with two other cases where the private sector was involved in the retention or transfer of data in a perspective of law enforcement:

56.

In the present EU PNR proposal, airlines have to make available in a systematic way the PNR data of all passengers. However, these data are not transferred directly in bulk to law enforcement authorities: they can be sent to an intermediary and they are assessed by a third party, the statute of which remains unclear, before selected information is sent to competent authorities.

57.

The main part of the processing happens in a grey zone, having material links with the first as well as with the third pillar. As it will be developed in Chapter IV, the quality of actors processing the data is not clear. Airlines are obviously no enforcement authorities, and intermediaries could be actors of the private sector. Even with regard to PIUs which would be public authorities, it must be stressed that not every public authority has the quality and the competences to perform law enforcement tasks on a routine basis.

58.

Traditionally, a clear separation has existed between law enforcement and private sector activities, where law enforcement tasks are performed by specifically dedicated authorities, in particular police forces, and private actors are solicited on a case by case basis to communicate personal data to these enforcement authorities. There is now a trend to impose cooperation for law enforcement purposes on private actors on a systematic basis, which raises the question which data protection framework (first or third pillar) applies to the conditions of this cooperation: should the rules be based on the quality of the data controller (private sector) or on the purpose followed (law enforcement)?

59.

The EDPS has already recalled the risk of a legal loophole between the first and third pillar activities (31). It is indeed far from clear whether activities by private companies, in some way connected with enforcement of criminal law, are covered by the field of action of the European Union legislator under the Articles 30, 31 and 34 TEU.

60.

If the general (first pillar) framework would not apply, a service provider should have to make difficult distinctions within his data bases. Under the current regime, it is clear that the data controller has to respect the same data protection vis-à-vis the data subjects irrespectively of the purposes that justify the retention of data. An outcome in which processing by service providers for different purposes would be subject to different frameworks for data protection should therefore be avoided.

61.

The different legal regimes that would apply at national level would have a major impact primarily on the exercise of his/her rights by the data subject.

62.

It is stated in the preamble of the proposal that ‘information, access, rectification, erasure and blocking, compensation and judicial remedies are to be provided under the data protection framework decision’. However, this statement does not answer the question of who the controller in charge of answering data subjects' requests is.

63.

While information on the processing could be communicated by airlines, the issue is more complex when it comes to access or rectification of data. These rights are indeed restricted under the data protection framework decision. As stated above, it is doubtful that a service provider such as an airline could be obliged to give differentiated access and rectification rights to the data it holds, depending on the purpose (commercial or law enforcement) followed. One might argue that these rights are to be exercised before the PIU or the otherwise designated competent authorities. The proposal however gives no further indication in this respect, and as already mentioned, it is not clear either that these authorities (at least PIUs) will be law enforcement authorities normally entrusted with restricted (possibly indirect) access procedures.

64.

The individual also risks being confronted with different recipients of data, as far as PIUs are concerned: the data are indeed transmitted to the PIU of the country of departure/arrival of flights, but also possibly to PIUs of other Member States on a case by case basis. Moreover, it is possible that several member States may establish or designate one single and common PIU. The data subject might in that case have to exercise redress before an authority of another Member State. Here again, it is not clear whether the national data protection rules will apply (these are supposed to be harmonised within the EU), or if specific law enforcement legislation will have to be taken into account (given the lack of comprehensive harmonisation in the third pillar at national level).

65.

The question is the same with regard to access to data processed by intermediaries, the statute of which is unclear, and which could also be common to airlines in different countries of the EU.

66.

The EDPS deplores the uncertainty that remains with regard to the exercise of these fundamental rights of the data subject. He stresses the fact that this situation is mostly due to the fact that actors who do not have law enforcement as a principal task are entrusted with such responsibilities.

67.

The EDPS considers that the proposal should make clear what legal regime is applicable at which stage of the processing, and specify vis-à-vis which actor or authority access and redress shall be exercised. The EDPS recalls that according to Article 30.1. b) TEU, provisions on data protection should be appropriate and cover the full range of processing operations established by the proposal. A simple reference to the data protection framework decision is not sufficient, given the limited scope of that framework decision and the restriction of rights it contains. As far as law enforcement authorities are involved, the rules of the data protection framework decision should at least apply to the whole processing foreseen in the proposal, in order to guarantee the coherence of the application of data protection principles.

68.

The EDPS notes that the proposal does not provide for any specification with regard to the quality of the recipients of personal data collected by airlines, be it for intermediaries, Passenger Information Units, or competent authorities. It must be stressed that the quality of the recipient is in direct relation with the type of data protection guarantees applying to that recipient. The difference between guarantees provided in particular by first and third pillar rules has already been mentioned. It is essential that the applicable regime be clear for all actors involved, including national governments, law enforcement agencies, the data protection authorities, as well as data controllers and data subjects involved.

69.

No indication is given in the proposal with regard to the quality of intermediaries (32). The role of intermediaries as controllers or processors is not specified either. From experience, it seems that a private sector entity, be it a Computer Reservation System or another entity, could perfectly be entrusted with the task of gathering PNR data directly from the airlines to redirect them to PIUs. It is indeed the way data are processed under the PNR Agreement with Canada. SITA (33) is the company responsible for the processing of the information. The role of the intermediary is decisive, as it could be responsible for the filtering out/reformatting of data that are transmitted in bulk by airlines (34). Even if intermediaries are obliged to cancel the processed information once it has been transferred to PIUs, the processing in itself is highly sensitive: a consequence of the intervention of intermediaries is the creation of an additional database including massive amounts of data, and even, according to the proposal, sensitive data (intermediaries being obliged then to delete those sensitive data). For these reasons, the EDPS recommends that no intermediaries should be involved in the processing of passenger data, unless their quality and tasks are strictly specified.

70.

PIUs have a decisive role in identifying persons that are or may be involved or associated with terrorism or organised crime. According to the proposal, they will be responsible for creating risk indicators and providing intelligence on travel patterns (35). Where the risk assessment is based on standardised travel patterns and not on material evidence linked with a concrete case, the analysis can be considered as constituting proactive investigation. The EDPS stresses that this kind of processing is in principle strictly regulated in Member State legislation (if not prohibited), and it is the task of specific public authorities the functioning of which is also strictly regulated.

71.

PIUs are therefore entrusted with very sensitive processing of information, without the proposal giving any detail on their quality and the conditions in which they would exercise this competence. Although it is likely that this task will be performed by a governmental body, possibly customs or border control, the proposal does not explicitly prevent Member States to entrust intelligence agencies or even any kind of processor with its performance. The EDPS underlines the fact that the transparency and guarantees applying to intelligence agencies are not always identical to those applicable to traditional law enforcement authorities. Details on the quality of PIUs are decisive, as this will have direct consequences on the applicable legal framework and the conditions of supervision. The EDPS considers that the proposal must include an additional provision detailing the specificities of PIUs.

72.

It appears from Article 4 of the proposal that any authority responsible for the prevention or combating of terrorist offences and organised crime can receive the data. While the purpose is clearly defined, the quality of the authority is missing. The proposal does not foresee any limitation of recipients to law enforcement authorities.

As mentioned above with regard to PIUs, it is decisive that the sensitive information at stake be processed in an environment with a clear legal framework. This is much more the case, e.g., for law enforcement authorities than for intelligence agencies. Considering the data mining elements and the proactive research included in the proposal, it cannot be excluded that such intelligence agencies be involved in the processing of the data, without exclusion of any other type of authorities.

73.

As a general comment, the EDPS notes that the enforcement of an EU PNR system is rendered even more difficult considering that law enforcement authorities have different competences depending on the national law of the Member States, including or not intelligence, tax, immigration or police. This is however a supplementary reason to recommend that the proposal be much more precise with regard to the quality of the mentioned actors and the guarantees to control the performance of their tasks. Additional provisions should be integrated in the proposal, to specify strictly the competences and the legal obligations of intermediaries, PIUs and other competent authorities.

74.

The proposal provides for some safeguards in relation to the transfer of PNR data to third countries (36). In particular, it foresees explicitly the application of the data protection framework decision to data transfers, it provides for a specific purpose limitation and it states the need for consent of the Member State in case of onward transfer. The transfer should also comply with national legislation of the Member State concerned, as well as any applicable international agreement.

75.

Many questions remain however, in particular with regard to the quality of consent, the conditions of application of the data protection framework decision and the question of ‘reciprocity’ in the transmission of data to third countries.

76.

The Member State of origin must give express consent for onward transfer of data from a third country to another third country. The proposal does not specify under what conditions and by whom this consent will be given, and whether national DPAs should be involved in the decision. The EDPS considers that the way consent will be given should at least be in conformity with national laws stating conditions of transfer of personal data to third countries.

77.

Besides, consent of a Member State should not prevail over the principle according to which an adequate level of protection must be foreseen by the recipient country for the intended processing. These conditions should be cumulative, as they are in the data protection framework decision (Article 14). The EDPS therefore suggests adding a point (c) to paragraph 1 of Article 8 that would read ‘and (c) the third State ensures an adequate level of protection for the intended data processing.’ The EDPS recalls in this respect that mechanisms ensuring common standards and coordinated decisions with regard to adequacy must be put in place (37).

78.

The proposal refers to the conditions and safeguards contained in the data protection framework decision while also specifying explicitly some conditions, in particular the above mentioned consent of the Member State concerned, and a limitation of the purpose to preventing and fighting terrorist offences and organised crime.

79.

The data protection framework decision itself provides for conditions to the transfer of personal data to third countries, namely with regard to the purpose limitation, the quality of recipients, the consent of Member State, and the adequacy principle. However, it also foresees derogations to these conditions of transfer: legitimate prevailing interests, especially important public interests, can be a sufficient basis for transfer even if the conditions listed above are not fulfilled.

80.

As already mentioned in Chapter III of this opinion, the EDPS considers that it must be stated clearly in the text of the proposal that the more precise guarantees of the proposal prevail over the general conditions — and exceptions — of the data protection framework decision, where it applies.

81.

The proposal addresses the issue of possible ‘retaliation requests’ of countries that might ask the EU for PNR data for flights from the EU to their territory. Where the EU requests data from databases of airlines of such third countries, because they operate a flight to or from the EU, such third country could ask the same from airlines based in the EU, including data from citizens of the EU. Although the Commission considers this eventuality as ‘very remote’, it allows for it. The proposal refers in this regard to the fact that the agreement with the US and with Canada foresee such reciprocal treatment ‘which may be enforced automatically’ (38). The EDPS questions the significance of such an automatic reciprocity and the application of safeguards to such transfers, notably taking into account the existence of an adequate level of protection of the country concerned.

82.

A distinction should be made between third countries which have already concluded an agreement with the EU, and those countries which do not have such agreement.

83.

The EDPS notes that reciprocity might lead to transfer personal data to countries where no guarantees can be provided in terms of democratic standards and adequate level of data protection.

84.

The impact assessment gives further elements with regard to the conditions of transfer of data to third countries: the advantage of the EU-PNR system, where data are filtered by PIUs, is emphasised. Only selected data of suspected individuals (and not bulk data) would be transferred to the competent authorities of Member States and presumably to third countries as well (39). The EDPS recommends clarifying this point in the text of the proposal. A simple statement in the impact assessment does not provide for the necessary protection.

85.

While the selection of data would contribute to minimise the impact on the privacy of passengers, it must be recalled that data protection principles go far beyond data minimisation, and include principles such as necessity, transparency and exercise of data subject rights, all principles to be taken into account when determining whether a third country provides for an adequate level of protection.

86.

The impact assessment indicates that such processing will provide the EU with the ability ‘to insist on certain standards and to ensure consistency in such bilateral agreements with third countries. It will also provide for the possibility of requesting reciprocal treatment from third countries with which the EU has an agreement, something that is not possible today’ (40).

87.

From these observations arises the question of the impact of the proposal on the existing agreements with Canada and the USA. The conditions of access to data in these agreements are indeed much broader, as they are not subject to a similar selection before being transferred to those third countries.

88.

The impact assessment indicates that ‘in cases in which the EU has an international agreement with a third country for the exchange/transmission of PNR data to such third country, such agreements shall be duly taken into account. The carriers should send the PNR data to the Passengers Information Units according to the normal practice under the current measure. The PIU which receives such data shall transmit them to the competent authority of the third country with which such an agreement exists’ (41).

89.

While on the one hand, the proposal seems to aim at a transfer of only selected data to any competent authority, be it within the EU or outside, on the other hand, the impact assessment, the preamble of the proposal (recital 21) and Article 11 itself recall that existing agreements should be duly taken into account. This might lead to the conclusion that filtering may only be a valid measure for agreements to be concluded in the future. It could be foreseen in this perspective that bulk access will still be the rule for access e.g. by US authorities to PNR data, in conformity with the provisions of the EU-US agreement, but that in parallel and on a case by case basis, a transfer of data to the US could occur, relating to specific data identified by PIUs, including but not limited to data concerning flights to the US.

90.

The EDPS regrets the lack of clarity on this decisive point of the proposal. He considers of the utmost importance that the conditions of transfer of PNR data to third countries be coherent and subject to a harmonised level of protection. Besides, for reasons of legal certainty, precisions with regard to the guarantees applying to the transfer of data should be included in the proposal itself and not only in the impact assessment, as it is the case now.

91.

The EDPS notes that the proposal explicitly excludes that enforcement actions be taken by the Passenger Information Units and the competent authorities of the Member Sates only by reason of the automated processing of PNR data or by reason of a person's race or ethnic origin, religious or philosophical belief, political opinion or sexual orientation (42).

92.

Such precision is welcome as it limits the risks of arbitrary measures against individuals. The EDPS however notes that its scope is limited to enforcement actions by PIUs or competent authorities. It does not exclude, in its present wording, the automated filtering of individuals according to standard profiles, nor does it prevent the automated constitution of lists of suspected persons and the taking of measures such as extended surveillance, as long as these measures are not considered as enforcement actions.

93.

The EDPS considers that the notion of enforcement actions is too vague, and that, as a principle, no decision should be taken with regard to individuals only by reason of the automated processing of their data (43). The EDPS recommends modifying the text accordingly.

94.

The proposal gives in Article 5.2 an important precision as it makes clear that no obligation is put on airlines to collect or retain additional data to those collected for the initial commercial purpose.

95.

Several aspects of the processing of these data still deserve further comment:

The EDPS supports the points made in the WP29-opinion in this respect.

96.

Air carriers established outside the EU are required to push data to PIUs or intermediaries as long as they possess the technical architecture to do it. If this is not the case they will have to permit the extraction of data through the pull method.

97.

Allowing for different methods of communication of data depending on the airlines concerned will only raise more difficulties with regard to the control of the compliance of transfer of PNR data with data protection rules. This risks as well to distort competition between EU and non-EU airlines.

98.

The EDPS recalls that the push method, allowing airlines to keep control on the quality of data transferred and the circumstances of transfers, is the only admissible method with regard to the proportionality of the processing. Besides, it must consist of an effective push, that is, the data should not be sent in bulk to an intermediary but filtered at the very first step of the processing. It is not admissible that non necessary data — and data not included in Annex 1 of the proposal — be sent to a third party, even if those data are to be deleted immediately by this third party.

99.

Article 9 of the proposal foresees a 5 years retention period of PNR data, with an additional 8 years period where data are to be kept in a ‘dormant’ database that will be accessible in restricted conditions.

100.

The EDPS questions the difference between the two types of data bases: it is questionable whether the dormant database constitutes a real archive, with different methods of storage and retrieval of data. Indeed, most of the conditions put to the access to the dormant database consist of security requirements that could be applicable to the ‘five years retention database’ as well.

101.

The total duration of storage — that is 13 years — is in any case excessive. It is justified in the impact assessment by the need to develop risk indicators and establish patterns of travel and behaviour (46), the efficiency of which deserves further demonstration. While it is obvious that data can be retained as long as necessary in a specific case as far as an investigation is ongoing, no justification can support the retention of data of all passengers in total absence of suspicion for 13 years.

102.

The EDPS further notes that this retention period is not supported by the answers of Member States to the questionnaire of the Commission, according to which the average duration of storage required would be 3,5 years (47).

103.

Moreover, the period of 13 years is comparable to the retention period of 15 years in the most recent agreement with the United States. The EDPS has always understood that this long retention period was only agreed upon because of strong pressure by the US Government to have a much longer period than 3.5 years, not because it was in any stage defended by the Council or the Commission. There is no reason to transpose such a compromise — that only has been justified as a necessary result of negotiations — to a legal instrument within the EU itself.

104.

The Committee of Member States established under Article 14 of the proposal will be competent with regard to security issues including protocol and encryption of PNR data, but also with regard to guidance for common general criteria, methods and practices related to risk assessment.

105.

Apart from these indications, the proposal does not include any element or criteria with regard to the concrete conditions and framework of the risk assessment process. The impact assessment mentions that the criteria will ultimately depend on intelligence held by each Member State, which is constantly evolving. The risk assessment is to be performed in absence of uniform standards of identification of suspects. The extent to which the Committee of Member States will be able to play a role in this regard thus appears questionable.

106.

The proposal details a series of security measures (48) to be taken by PIUs, intermediaries and other competent authorities in order to protect the data. Considering the importance of the data base and the sensitivity of the processing, the EDPS considers that in addition to the measures envisaged, the entity processing the data should also be obliged to officially notify any security breach.

107.

The EDPS is aware of the project to establish such a notification procedure in the sector of electronic communications at European level. He advises to include such safeguard in the present proposal, and refers in this respect to the security breach system put in place in the United States with regard to state agencies (49). Security incidents can indeed happen in any field of activity, and in the private as well as the public sector, as the recent loss of a whole citizens' database by the British administration has shown (50). Large scale databases such as the one envisaged in the proposal would be first on a priority list to benefit from such an alert system.

108.

The EDPS notes that a review is to take place within three years of the entry into force of the framework decision, on the basis of a report prepared by the Commission. He acknowledges the fact that this review, based on information provided by Member States, will give specific attention to data protection safeguards, and include the implementation of the ‘push’ method, the data retention and the quality of the risk assessment. Such review should, to be comprehensive, include the results of an analysis of the statistical data produced on the basis of the processing of PNR information. These statistics should, in addition to the elements mentioned in Article 18 of the proposal, include statistical details on the identification of high risk persons, such as the criteria for such identification and the concrete results of any law enforcement action resulting of the identification.

109.

The EDPS has already insisted in this opinion on the absence of concrete elements to establish the necessity of the system proposed. He considers however that, should the framework decision enter into force, it should as a minimum be complemented by a sunset clause. At the end of the three years period, the Framework Decision should be repealed in case no element would come to support its continuation.

110.

In its final provisions the proposal puts a condition to the further application of already existing bilateral or multilateral agreements or arrangements. Those instruments can only be applied as far as they are compatible with the objectives of the proposed framework decision.

111.

The EDPS questions the scope of this provision. As already mentioned in Chapter V under Reciprocity, it is not clear what the impact of this provision will be on the content of agreements with third countries, such as the agreement with the US. In a different perspective, it is not clear either whether the provision could have an impact on the conditions of application of instruments with a broader scope, such as Council of Europe Convention No 108. Although this might appear unlikely in view of the difference of institutional context and actors involved, any risk of misinterpretation should be avoided and the proposal should make clear that it does not have any impact on instruments with a broader scope, notably those having as object the protection of fundamental rights.

112.

The EDPS stresses the major impact in terms of data protection of the present proposal. He has concentrated his analysis on four fundamental issues raised by the proposal, and insists on the fact that the issues raised need to be addressed in a comprehensive manner. Under the present circumstances, the proposal is not in conformity with fundamental rights, notably Article 8 of the Charter of the Fundamental Rights of the Union, and should not be adopted.

113.

Should the comments above be complied with, especially the legitimacy test, some drafting proposals have been made in the present opinion that should be taken into account by the legislator. Reference is made in particular to points 67, 73, 77, 80, 90, 93, 106, 109 and 111 of the opinion.

114.

While the general purpose to fight against terrorism and organised crime is in itself clear and legitimate, the core of the processing to be put in place is not sufficiently circumscribed and justified.

115.

The EDPS considers that techniques consisting of assessing the risk presented by individuals using data mining tools and behavioural patterns need to be further assessed, and their utility be clearly established in the framework of the fight against terrorism, before they are used on such a wide scale.

116.

Building upon different data bases without a global view on the concrete results and shortcomings:

117.

The fight against terrorism can certainly be a legitimate ground to apply exceptions to the fundamental rights to privacy and data protection. However, to be valid, the necessity of the intrusion must be supported by clear and undeniable elements, and the proportionality of the processing must be demonstrated. This is all the more required in case of extensive intrusion in the privacy of individuals, as foreseen in the proposal.

118.

These elements of justification are missing in the proposal and the necessity and proportionality tests are not fulfilled.

119.

The EDPS insists on the essential character of the necessity and proportionality tests developed above. They represent a condicio sine qua non to the entry into force of the proposal.

120.

The EDPS notes a serious lack of legal certainty as to the regime applicable to the different actors involved in the project, and in particular to airlines and other first pillar actors: be it the rules of the proposal, the rules of the data protection framework decision or the national legislation implementing Directive 95/46/EC. The legislator should make clear at what stages of the processing these different rules will apply.

121.

The present trend to impose cooperation for law enforcement purposes on private actors on a systematic basis raises the question which data protection framework (first or third pillar) applies to the conditions of this cooperation: it is not clear whether the rules should be based on the quality of the data controller (private sector) or on the purpose followed (law enforcement).

122.

The EDPS has already stressed the risk of a legal loophole between the first and third pillar activities (51). It is indeed far from clear whether activities by private companies, in some way connected with enforcement of criminal law, are covered by the field of action of the European Union legislator under the Articles 30, 31 and 34 TEU.

123.

An outcome in which processing by service providers for different purposes would be subject to different frameworks for data protection should be avoided, especially considering the difficulties this would create in terms of exercise of rights by data subjects.

124.

The proposal should provide for a specification with regard to the quality of the recipients of personal data collected by airlines, be it for intermediaries, Passenger Information Units, or competent authorities.

125.

The quality of the recipient, that could in some cases be private sector actors, is in direct relation with the type of data protection guarantees applying to that recipient. It is essential that the applicable regime be clear for all actors involved, including the legislator, the data protection authorities, as well as data controllers and data subjects involved.

126.

The EDPS stresses the need to ensure that an adequate level of protection is provided in the recipient country. He also questions the significance of the ‘reciprocity’ principle mentioned in the proposal, and its application to countries already bound by an agreement with the EU, like Canada or the US. He considers it to be of the utmost importance that the conditions of transfer of PNR data to third countries be coherent and subject to a harmonised level of protection.

127.

The EDPS also draws the attention of the legislator to specific aspects of the proposal that need more precision or a better taking into account of data protection principle. This is the case in particular with regard the following aspects:

128.

The EDPS notes that the present proposal is made at a moment when the institutional context of the European Union is about to change fundamentally. The consequences of the Lisbon Treaty in terms of decision making will be fundamental, especially with regard to the role of the Parliament.

129.

Considering the unprecedented impact of the proposal in terms of fundamental rights, the EDPS advises not to adopt it under the present Treaty Framework, but to ensure it follows the co-decision procedure foreseen by the new Treaty. This would strengthen the legal grounds on which the decisive measures envisaged in the proposal would be taken.

1.

protecting State security, public safety, the monetary interests of the State or the suppression of criminal offences;

2.

protecting the data subject or the rights and freedoms of others.’

1.5.2008   

EN

Official Journal of the European Union

C 110/16

1.5.2008   

EN

Official Journal of the European Union

C 110/17

—

from the Europa competition web site (http://ec.europa.eu/comm/competition/mergers/cases/). This web site provides various facilities to help locate individual merger decisions, including company, case number, date and sectoral indexes,

—

in electronic form on the EUR-Lex website under document number 32008M5084. EUR-Lex is the on-line access to European law (http://eur-lex.europa.eu).

1.5.2008   

EN

Official Journal of the European Union

C 110/17

—

from the Europa competition web site (http://ec.europa.eu/comm/competition/mergers/cases/). This web site provides various facilities to help locate individual merger decisions, including company, case number, date and sectoral indexes,

—

in electronic form on the EUR-Lex website under document number 32008M5104. EUR-Lex is the on-line access to European law (http://eur-lex.europa.eu).

1.5.2008   

EN

Official Journal of the European Union

C 110/18

—

from the Europa competition web site (http://ec.europa.eu/comm/competition/mergers/cases/). This web site provides various facilities to help locate individual merger decisions, including company, case number, date and sectoral indexes,

—

in electronic form on the EUR-Lex website under document number 32008M4784. EUR-Lex is the on-line access to European law (http://eur-lex.europa.eu).

1.5.2008   

EN

Official Journal of the European Union

C 110/18

—

from the Europa competition web site (http://ec.europa.eu/comm/competition/mergers/cases/). This web site provides various facilities to help locate individual merger decisions, including company, case number, date and sectoral indexes,

—

in electronic form on the EUR-Lex website under document number 32008M5092. EUR-Lex is the on-rline access to European law (http://eur-lex.europa.eu).

1.5.2008   

EN

Official Journal of the European Union

C 110/19

—

from the Europa competition web site (http://ec.europa.eu/comm/competition/mergers/cases/). This web site provides various facilities to help locate individual merger decisions, including company, case number, date and sectoral indexes,

—

in electronic form on the EUR-Lex website under document number 32008M5061. EUR-Lex is the on-line access to European law (http://eur-lex.europa.eu).

1.5.2008   

EN

Official Journal of the European Union

C 110/20

1.5.2008   

EN

Official Journal of the European Union

C 110/21

1.5.2008   

EN

Official Journal of the European Union

C 110/22

Gobierno de Navarra

Departamento de Desarrollo Rural y Medio Ambiente

C/ Tudela, 20

E-31003 Pamplona (Navarra)

Dirección General de Desarrollo Rural

C/ Tudela, 20

E-31003 Pamplona

Tlf: (34-848) 42 29 33

E-mail: izabalzv@cfnavarra.es

Gobierno de Navarra

Departement Landbouw en Visserij

Duurzame Landbouwontwikkeling

Ellips, 6e verdieping

Koning Albert II laan 35, bus 40

B-1030 Brussel

—

Article 4 of Regulation (EC) No 1857/2006

—

Code général des collectivités territoriales (partie législative): articles L 1511-1 et L 1511-2

—

Délibération de la commission permanente régionale du 12 octobre 2007

(a)

the construction, acquisition or improvement of immovable property;

(b)

the purchase or lease-purchase of new machinery and equipment, including computer software up to the market value of the asset;

(c)

general costs linked to expenditure under points (a) and (b), such as architects, engineers and consultation fees, feasibility studies, the acquisition of patents and licences

Conseil régional du Centre

9, rue St Pierre Lentin

F-45041 Orléans cedex 1

Tél.: (33) 02 38 70 30 30

Fax: (33) 02 38 70 31 18

E-mail: Amanda.Tibble@regioncentre.fr

Departamento de Agricultura, Alimentación y Acción Rural

Generalitat de Catalunya

Gran Via de les Corts Catalanes, 612-614

E-08007 Barcelona

2007: EUR 70 000

2008: EUR 66 750

2009: EUR 67 500

2010: EUR 67 875

2011: EUR 68 250

2012: EUR 68 625

2013: EUR 69 000

—

up to 50 % of eligible costs in less-favoured areas and up to 40 % of eligible costs in other areas.

—

for investment in non-productive features, up to 100 % of actual costs,

—

for investment in productive assets on farms, up to 60 % of actual costs, or 75 % in less-favoured areas, provided that the investment does not entail any increase in the production capacity of the farm,

—

additional aid may be granted at a rate of up to 100 % to cover the extra costs incurred by using traditional materials necessary to maintain the heritage features of buildings.

—

the amount of municipal co-financing is the difference between the amount of co-financing of insurance premiums from the national budget and up to 50 % of eligible costs of an insurance premium for insuring crops and fruit and insuring livestock against disease.

—

up to 100 % of actual legal and administrative costs incurred.

—

up to 100 % of actual costs incurred; this is to be provided in the form of subsidised services and must not involve direct payments of money to producers.

—

up to 100 % of eligible costs concerning farmers' education and training, consultancy services, the organisation of forums, competitions, exhibitions, fairs, publications and catalogues, the dissemination of scientific knowledge and the costs of replacement services. The aid is to be granted in the form of subsidised services and must not involve direct payments of money to producers

Občina Žužemberk

Grajski trg 33

SLO-8360 Žužemberk

1.5.2008   

EN

Official Journal of the European Union

C 110/27

1.5.2008   

EN

Official Journal of the European Union

C 110/28

Block C, Sector 45

Greenwood City

Gurgaon 122003

Haryana

India

Express Building 9-10

Bahadur Shah Zaraf Marg

New Delhi 110002

India

1.5.2008   

EN

Official Journal of the European Union

C 110/29