 Okay Good morning, everyone. Bon dia. And now it's time for you. So The idea is that first of all I would like to to invite the speakers if they want to Add some comments to other people speak talk with I don't know if Andreas wants to add something About what has been told before I afterwards with No, because because you mentioned something I have some questions, but before if you want if you want to talk about other other the other speakers presentation because for instance when when you present the the study it was somehow understood that everything has to be open and It has been changed a little bit because it's what Jean-Claude has mentioned about the open by default, but as open as possible as close as necessary So maybe then we could fulfill I have another question about legal question if It could be legal to deposit Data Closely there is a project from Harvard called data tax that they are offering tools to deposit data even with Encrypted data when there is sensitive data and then at least Everybody could have like a catalog of data knowing that the data exists even the data is closed So maybe then we can find some protocols So that's one question for you another question for Christophe for instance I I agree with some comments about the public domain mark and the creative commons license I don't I know that in the afternoon. There is a Breakout session dedicated to license so we can discuss further more about the data and the problem that suppose in Europe the database Rights that also has another issue and to Jean-Claude. I also have another question about Who evaluates the opt out so meaning I can say I have IPR issues I have but do you evaluate that or you just? Accept it as it is so I don't know just three questions that I just jump and I will invite the other people to Yeah, as you addressed me. Maybe I'll start Yeah, it was very interesting to to hear that it's not about openness as a utmost principle but as Open as possible But of course considering what I mentioned about data protection We have a here basic conflict and the question is if we Really adhere to what data protection? May force us to do is it still open access in the end what we come out of it so I'm not sure if it's really possible to Absolutely Live up to data protection standards and still keep the level of open access as we want to of course It's always a way to find solutions and there are solutions and one solution may be encryption you mentioned. This is the other Next to anonymization may be the other technical way out of the problems and it's also just discussed on a general level As a means for protection of data in general to have it encrypted on a very low technical basis So as soon as data produced that will be encrypted and only decrypted at the end user or the user who really used it That I'm not an informatic But what I read and heard from informatic people it may be a way But of course then you create another problem of management because as you mentioned you have to give people Access to the encrypted data and then you have to draft again some kind of management system which also includes Giving certain people rights and other people not rights and so and create new problems of management of such a system. So Everything has its ups and downs and it's it's very difficult to but maybe Niels also has some something to add he didn't Involved yet, so I think you also has something to say Very challenging technique well, I would Have a question for John Claude You Who showed us the statistic of the opt-outs out of the pilot my question is Can't I as a project participate in the pilot and write a data management plan that says I Don't open up any data Isn't this possible? So is the statistic somehow not a whole reality It's my thing So So on to To take the two questions maybe at the same time so So I would I would find it extremely remarkable if if a project opting in or opting out in 2016-2017 of post Christus These days will not think about a DMP plan full stop I mean in our data-driven science reality It's a bit remarkable if people would still Consider submitting large consortia large projects without thinking of DMP Regardless if they opt in or out I think this should be standard and I actually think doctoral school should give a mandatory Course course on that to be more precise on your question Of course all the opt-out projects can certainly Write the DMP and and it is actually encouraged also somewhere in the rules But it is not taken into consideration in the in the in the evaluation So it is not used as as an argument later on if the if the project complied with all the requirements So we do encourage that projects write DMP, but if they opt out they don't share it You see so that is because that they don't want to I mean they share the data to the project officers and within the construction But I don't share it in repository. So that is it is it is in a certain sense Zero-sum game in that respect and and as you said so who evaluates this well do we have All our project officers are to the people organizing the evaluation and Bringing together the panels for evaluating the projects they they they do they do the job there And I also think if I made to To come back on what you said I Really think we have to try to come up with a dynamic approach into defining these these problematic issues of personal and Privacy and so on because a static a static approach will Simply simply not work if if if in ten years time we all have implants to regulate something Now before we fix that legally. I mean Pharmaceutical industry will have already done it. So so it is it is something we have to say Probably more on a on a case by case base and the cases can be enormous They are not Mickey Mouse cases here and we should try to come up with the dynamic approach And that's why I think and that's what the point I wanted to make I Think it's very important that the the community and the disciplines themselves. I have an important saying in this that it is not only let Not only the the the mandate of legal or Technical people to judge on this but communities themselves should should have a voice in in this in the in deciding on what the dynamic Framework could be Yes, well actually I would like to since you mentioned the the license Question again Picking the license is a bit tricky The and maybe And they asked we can comment on it because they are he's the he's the real expert on that The CC community claims that using the CC zero License is possible even Within jurisdictions that limit what you can wave There has even been a Ported version just a I don't know a month or two ago of the new of the CC by of the CC suit 4.0 you know transferred to to German Not only the language but changed a little bit. I'm not a hundred percent sure if the Lawyers would fully agree That the CC zero license is fully applicable That's that's the problem so if it's if people feel comfortable they can use it beside this This lead this possible legal issue. There's this psychological thing to which I alluded earlier For many people it is very difficult to wave Alright, so it may be something we like but that actually brings me to what I consider my main message. I think We need a framework that is to be provided by the research Organizations that takes a lot of the responsibility concerning making data available away from the researcher and provides Certain workflows which you are then perceived as just being the normal workflow and then it just goes that way If we ask the people to do all these individual data management plans and everything I mean that they may be applicable on the strictly scientific Level but concerning the legal level. It's far too complicated for the scientists. We need to provide for them a Set of rules that works kind of across the board as long as they are a member of a certain organization Yeah, as to the waving of rights the CC with the CC license Of course the situation in different countries is different as well in this respect You also have to take into account that this different rights involved is not only copyrights Also, the database right may be involved which is has come quite late into the CC complex as well and For copyright for example in our continental system. It's not possible to wave copyright completely Whereas the database right is the situation a bit different the problem is But that maybe two more problems also with this waving of or with this CC zero right which I Think applies when there are no rights involved The first the first one is if you if you have a license that is not based on any right copyright or other right It only is valid between the parties So once this the information gets out outside of the contract to a third party You cannot control it anymore this kind of license unless the third party is also bound by this license and of course Even if there is no right copyright or other rights involved the validity of Such a license is still subject to general civil law standards and this means for one thing. There's a Kind of standard control Mechanism which is has different levels of strictness in different countries for example generally have very strict control of standard contracts is also European directive, but the the levels in the member states are different of this kind of control and there's also a View review according to competition law cartel law and these are the main subjects There is a you for example European court decision if if you license for example database that is not protected by database rights you can Put as many Restrictions as possible but subject to national law and then you have to look at the national law German Italian French and so on How strict the control of these contracts arms? That's the problem that we have different levels even within Europe of this kind of control. So it's a complicated situation as well Perfect you are preparing the session for at the afternoon Please if there is any question there are a few questions there So I pass around and you can introduce yourself, and then we know you come from Your guy from Germany. I'm in the bio banking field, and so we are very much concerned with privacy questions What we came basically that so basically one comment and one question One comment would be what we came down basically is with regard to data quality. There's no way to get off the requirement for consent because Anonymization basically means a reduction of data quality to a to a level that they are no longer usable in a really Reasonable way so we have to stick to a consent and our aim is to implement kinds of a broad consent at which is Basically, it's a legal problem actually that as long as we are and that's our opinion as long We are flanking methods to that like ethical committees IRB's technical restrictions We hope that we can implement it in a way that it's usable and it's acceptable from the legal and From the individual I think this is a way to go Because there are no real alternatives in my opinion to that And just a question following that Different issue, but also legal that's the question of responsibility what we are very much concerned as well in the medical field is that whenever you give data out there's someone using the data and How can we control well? How can we be responsible for data that have been generated somewhere and Are now used for clinical decisions for example in the early early clinical a unique for example If the data are not valid or of very low quality there are consequences there may be consequences for the patient There may be consequences in a very broad range Who takes the responsibility if these are open data? And we actually cannot stand for the quality to for the data because we haven't generated them We only have collected them so so maybe I I start with You your first comment I I agree's consent is as of now the the main legal basis to work with and There as I mentioned we still have a problem that It's a bit uncertain under which Which qualifications consent may be valid even to national law data protection laws? I think it would be helpful if the data Protection group 29 on all its follow-up another regulation would make some general statement on Like a commentary on what could be the valid requirements for a valid consent and to Introduce this into the legal discussion. Hopefully we will see this in the future The last question is quite complicated goes goes back to his general civil law. Is there like any kind of liability for? information quality as of now the Standards for liability for information are quite broad so Cannot really make somebody liable for bad bad information quality, of course there are limits to everything but Generally under tort law at least it's it's difficult to make somebody liable. The other question is I'm not sure if That was also your direction if you Produced personal data and give it to somebody else are you liable what somebody else is doing with the data and the data protection law You know the liability is limited to the data processors as a person who processes the data Of course if I give data to somebody else under the new regulation, I'm and To a certain degree also responsible in the sense that if the data subject requires me to erase the data and I gave this data also somebody else a third party. I have to use efforts to also get him to erase the data I'm not I'm not complete responsible, but I have to provide efforts to to do this So there's some kind of data protection also Responsibility for what third parties are doing with data. I gave to them, but maybe Niels also has to add Well, it depends on what the well new the data processor is doing with the data. So Yeah, okay, we have another question here Hi, my name is Meda Devare and I'm an agricultural scientist I work at CGIR, which is a group of 15 centers where we mostly are located in developing countries and And our mandate is to reduce poverty to to increase productivity reduce, you know increase food security And natural resource management that sort of thing. We're about 10,000 staff members mostly scientists So, you know, this was very interesting. It's raised at least three issues for me three major issues What one is the the the issue of of Well data as you know the idea of ownership I think you mentioned that that data cannot be owned but increasingly what we're finding is that in the countries we work Depending on which country it is civil society organizations, particularly are seeing data as as a monetary Asset and so that becomes a real issue for us as we work, you know with with smallholder farmers So that's that's a question I guess that I that I would throw to you and ask for you know if you have any thoughts on that The second issue is I think what was raised earlier the issue of consent when we're working with smallholder farmers who are primarily Illiterate or semi-literate. It's a very hard thing to explain You know, how do we deal with this issue of consent and and who benefits and and and the idea of benefit sharing I mean for us I think our assumption is the benefit sharing takes the form mostly of Being able to come back to the farmer or the farmers the community with improved ways of doing their business Which is you know improved productivity hopefully translates to improved profitability and and those are the assumptions behind what we do Doesn't always play out and we are taking the data out. So so there are those sorts of thorny issues to deal with So so how would you how would you? You know what what sort of guidance would you would you provide and especially as funding agencies? How do you how do you reconcile what you're requiring? With with what's the reality on the ground? I mean, I know that that with the EC I'd like to that statement about you know as open as possible as closed as needed But our primary funders often don't allow that kind of thing You know, there's a sort of a bottom line thou shalt make it open and while I believe in that It's very hard to actually follow and practice So there there are all kinds of thorny issues when you're working not in Europe not in the US You know not in North America. How what what sorts of guidance would you give for that sort of situation? Okay, so Works so let me try to Give a few elements because again, I think this is an unfolding discussion for which No one has a Certain answer I would say otherwise you will not sit here. We just do it No, I think I think The requirement for open access is is unstoppable and you know, we we we have Power points Floating around in in Brussels where we give one example after the other but if if if big private funders like welcome and the Gates Foundation Go for mandatory Open access across across the lifecycle almost of the research process I think it says something and and and public funders are are slowly following the Commission wants to to lead by example here So we try to to to speed up the process and Because it is there are a lot of emotions in this discussion and and there is not always very rational arguments there in In in all this because to a certain to a certain degree consent ownership These issues have been on the table as long as science existed Except that now it is it is multiplied at the at scale X because of the digital Availability and the internet so so it is these are not issues that we have not looked at in the past and quality of data and Quality of data production and the reuse of data or the malign reuse of data These are things that have been existed all the time and for which there were Sometimes Solutions the complexity becomes because of of the fact that it is now available at the scale which was unprecedented and I personally also think of for example if you if you look at the issue of of of Ownership well if you If you try to restrict the question to a research environment, it becomes easier to tackle Because ownership of data in general I cannot see how you can solve that in in general because it is simply too fast and too dynamic, but if you Refind the question to Who owns the data that have led to certain pieces of research which have been Pre-reviewed and accepted by the community community. It becomes much easier because you know that there is quality Otherwise, it would have been peer-reviewed. You know who is the author. You know, you know that there are probably Instruments being put in place for for for consent. So I think that that is really something we have to See at different levels. I the example I always use is that hospitals are very Easily to be convinced to to to live to give their data if it is for research question to find for example how Asthma is related to traffic jams in in urban cities They will not do so if the project is sponsored by a pharmaceutical company or they will do some condition that The IPRs are taken into account and they will certainly not do so if they don't know where the question comes from So there are gradations in the problem there which which are related to the nature of the research question, but a Generic open data policy is simply impossible now and if I if I made to to To tackle the issue of quality I think if we go but that's a speculation if we go to an environment where open data becomes a default attitude of a researcher Which means that the downloadability of his data might be an indicator of the quality of his of him as a scientist as a data Provider and that will be and that can be then taken into account in the evaluation for his career for his tenure And you can easily imagine in 20 years time that how many peer-reviewed Artics did you publish how many data sets are and what what has been used of that as an example now That is an inherent mechanism of the science world that we know works very well, which is peer-reviewed Controls so you can assume that the more open the system is the better the quality will become because there is much more control So that is a little bit the way I would I would look for For for solutions and then finally on on the consenting and I think it's it's very important because we will see much more Much that's my personal view We will see more and more Distributed ways of collecting data rather than centralizing them You we all know the example of the health applications on your iPhone that at some point in time It was discovered that actually and in the back of your activity the health application Registered how many steps you did how many kilometers you run and then those data were sold or used by by by Apple related companies to do XYZ on that so that kind of mechanism whereby the data remains on your phone But the the software so to speak to use them is traveling around is a completely new challenge because if it is All centralized in in one data center, then it's relatively easy if it is no longer Centralized but decent actually a little bit blockchain if you want you can even think in that direction in the future Then then you get real big tough Questions on the table now. How do we organize consent here? And what are the the procedures to put in place? You mentioned the ownership question again I think it's also this question is now so much discussed because of the Increasing amount of data production and data processing and we're talking about internet of things now We will in 10 years will be all be part of like a thing in the internet of things if you look Here what informatics are projecting. It's a bit scary. Also. I must say The owner the ownership question and I know very well in practice People think that data they have stored are their property and you can also look into contracts or you can look into standard terms We provide you with the ownership of the data or we transfer the ownership of it from a legal point of view There is no ownership in data. There's no property in data What what you might have is in very exceptional cases copyright on on parts of the information What you might have is a database, right? the database right if you extensively Constru it may apply to censored To data which are produced by sensors and if you from the sensor it goes directly into database you might include this Data produced by sensors into the database, right? But to the rest of course, yes trade secret, but this doesn't play much role in research field Beyond that. There's no property right in data Of course, the problem is the practice is not aware of it And even if you go to the industry people don't know this and though they treat data as if they're their property That's why also you've been commissioner thinking about introducing array right which would accrue to the producer of the data But there are lots of practical problems. For example, I always take the example of a car The cars we use today have about 100 sensors Included already they're producing all kinds of data You can measure the air in the car and to evaluate if the driver is drunk, for example, it's possible already today So then if you have this data, you know about how the car is driving the speed what the driver is doing Who should be the owner of this data? Is it the driver? Is it the owner? Is it the producer of the car? Is it the producer of the box with the sensors? All all our candidates for such a property right So who should be the owner you have to make a decision who will be the owner and then he can negotiate his property So already the question who should be the owner is quite unsolvable and the next problem is of course with personal data You can always say which person is the information about he's a data subject If you apply such a right beyond personal data to any kind of data Then you have to decide what is the subject of this right You mentioned data that are stored on my mobile if I like pictures or like Like address data or something if I transfer this data to another mobile Is it the same data or is it different data that are stored on the next mobile? I think a different data. So what is data about basically it's about information So you would have a right on information and then of course you have the same the problem I mentioned you would protect any information would be complete shift from freedom free flow of information to a Property concept of information and I think it wouldn't work in practice. So that's the problem I have with the property they General ownership of of course you have different levels of legal ownership. You can say one thing is property right You can also exclude others from using information a Bit lower level would be to just have Like defensive rights. So if somebody destroys my data, I can have a claim to damages But of course not the same as property right that this the second we have already So if somebody destroys my computer, I may also have a claim to the value of the information of the data That were destroyed, but it's not a right to control We are third parties So the question of data ownership is very interesting, but as I meant as I said, I think There will be in or even the future there will be no property right in data in general, but but only the Commission as least the policy of the Commission is more to To promote the free flow of information and if some kind of right is is good for promoting this free Flow of information we might might see it, but we won't have a property right in data in general And of course the question of consent. I think we we mentioned a couple of times already on the basis of existing laws very difficult to really have a Working practically working Concept of consent that really lives up to the data protection requirements What's what's happening the practice is some some kind of consent that is not really challenged in court So it's it's it's somehow is working practice, but if you really take the law seriously Most of what's happening practice is is not really according to I would say going to the law Okay Before the coffee because everybody's waiting for the coffee a final question Hi, I'm Prodromos Chiavos. I'm from the Athenae reset Center and also I'm serving at the European Patents Office Patent Academy so I'm I was wondering about the link between the patent data and the resets data and in in the diagrams you've shown We've seen that IPR is the primary reason for opting out But I think it is important to understand what exactly that means Because even if we see the patent system the patent system we tend to think about it only as a protection and as a closing system but it has a very strong component of disclosure and publication and The European Patent Office has an enormous wealth of data related to patents so My first question is to what extent the question of opting out is a timing question Rather than a substantial question of not releasing the data because it could be that I release the data Along or after I actually submit my patents and the second question is how could we align more the Data we have with regards to the state of the arts Which exists even either in the national patent offices or at the European Patent Office with the data that we have as every as research communities and What I mean is first to do a simple mapping between let's say Frascati and IPC and Secondly, how we could we actually try to reuse some of these data because they are exceptionally relevant For our researchers in order to position their research within the state of the art. Thank you I assume so On your first your first point Most of what you said I use as an answer when we are visitors from industry in our office. Yeah, but you know all this openness They they all they seem to think that this is that we want to install the last communist regime in Western Europe So it's a but hang on. I mean also before the Internet existed You could there was a way to protect your research and that was called patents and the patents I as you say are also a very good example in of Control disclosure, so I think that Again, that's why I think some of the concern is a little bit overblown And there are the what has worked in the past will also work in the digital age and that is patents is a fantastic Instrument to protect your your RPRs now on on the beginning of your question So the the box is ticked We want to opt out because of we assume that there might be a problem from an IPR point of view That's that's that's how it works and then what we now would like to see is a little bit more Substantiation of that But there again, I mean we what we try to win to to train our project officers now We know some but hang on if that is put on the table in the discussions Please flag To the to these good people that there still is something in like a patent that can That can do the job perfectly well, so you can I would say maybe not so much in health research per se but to a very large degree you can marry openness with with A property control as well if you use patents now and I personally think that it would be fantastic if the databases of the IPO would be used in in the science club so I can see you as a as a big provider there as one of the Notes for that kind of data in in the overall architecture of the science But you know the way we we see the science club It's it's a federation of data centers and and clouds and that that could be One of them now on the second question. I I lost it in the meantime, so what was the second question? That we have a research So I think this is more like a call to actually do it possibly in one of the subsequent projects How we could actually map? Classifications because they help the positioning of research within the state of the art as understood by the patent system I think that would be very valuable Yeah, but that that's why we have RDA know so that to find solutions for this kind of synchronizing the The data series indeed Yeah, I still remember the time when you went wanted to find out about patents You had to go to the library and there was that's the wall of books patent Patent Applikity had to browse through today is much easier already to the databases we have at hand Maybe just one small comment to the first question Patent doesn't protect the information But only the invention so the information is free So in this sense, there's no conflict between patent law and and free use of data for research unless there's some copyright sometimes there's come some copyright of course on on the patent file the patent application, but That's not really impediment to to using of the of the patent information. That's just the idea of the patent system to Yeah Yeah, okay. Okay in the early stage Yeah, okay, I want to add something not to this question, but to another question with the Ownership of data. This is also a very practical problem with the use of the CC licenses for example, so often Data which is not protected by IP rights is licensed for example by the CC by Which actually doesn't make a sense if there is no copyright protection, but it then has Kind of a factual protection because the next user is using this license again, so it's Yeah, it appears this protection which is actually not existing Okay, I completely agree with you in fact We are making a document a lot of people from CC and that was one of our points to please Don't use CC by when there is no copyright. So thank you very much for this panel