“Data is pow­er” – Is this a view tak­en by the antitrust author­i­ty when look­ing at Big Data pools?

The sig­nif­i­cance of data for com­pe­ti­tion is grow­ing, as com­pa­nies can use it to opti­mise their own pro­duc­tion process­es, sales struc­tures and mar­ket­ing cam­paigns in com­pe­ti­tion with oth­er com­pa­nies, per­son­alise ser­vices, iden­ti­fy mar­ket trends at an ear­ly stage or shape them them­selves. Every com­pa­ny needs data­bas­es, data pools, data lakes or data ware­hous­es for this pur­pose. When does this con­sti­tute abu­sive behav­iour? When does a com­pa­ny have to pro­vide access to “its” data?

In the lat­est of her Duet inter­views, Dr Cal­daro­la, edi­tor of Data Ware­house as well as author of Big Data and Law, and Dr Grace Nacimien­to dis­cuss  the com­pe­ti­tion aspects of data warehouses.

I can well imag­ine that every com­pa­ny has or needs data pools, data lakes or data ware­hous­es if it wants to par­tic­i­pate in the process of dig­i­tal­i­sa­tion. This means that not only do they have data, ‑but the infor­ma­tion and knowl­edge gained from it as well, which gives them a com­pet­i­tive advan­tage and lim­its com­peti­tors in the mar­ket. In the book “Data Ware­house”, you exam­ined the antitrust law aspect. Is this fact or are these cir­cum­stances rel­e­vant under antitrust law and why?

Dr Grace Nacimien­to: Your ques­tion goes right to the heart of data and anti-trust laws. Data has become a key fac­tor for com­pa­nies’ com­pet­i­tive­ness in the dig­i­tal econ­o­my. Com­pa­nies that col­lect and analyse large amounts of data often have a sig­nif­i­cant com­pet­i­tive advan­tage. This can lead to a dom­i­nant mar­ket posi­tion, which is rel­e­vant under com­pe­ti­tion law. So, you can in fact say that data is pow­er and there­by has sig­nif­i­cant poten­tial for influ­enc­ing mar­ket entries, mar­ket posi­tions and mar­ket con­di­tions. Gain­ing access to and con­trol over large amounts of data can cause restric­tions of com­pe­ti­tion. If a com­pa­ny denies or restricts access to data, this can sig­nif­i­cant­ly affect com­pe­ti­tion in the mar­ket. This is par­tic­u­lar­ly rel­e­vant if the data is cru­cial for devel­op­ing new prod­ucts or ser­vices. Car man­u­fac­tur­ers, for exam­ple, have the nec­es­sary data for main­te­nance and repair of their vehi­cles, enabling them to influ­ence who will be able to offer main­te­nance and repair ser­vices on  the mar­ket. Dig­i­tal com­pa­nies col­lect per­son­al data of their end users which can be used to mar­ket addi­tion­al ser­vices and prod­ucts. Patient and pre­scrip­tion data col­lect­ed in phar­ma­cies could be used by phar­ma­ceu­ti­cal com­pa­nies or con­sul­tan­cy firms to devel­op mar­ket­ing strate­gies grant­i­ng com­pet­i­tive advan­tages to those who have access to the rel­e­vant data – of course, in con­for­mance with data pro­tec­tion require­ments when deal­ing with per­son­al and sen­si­tive data. The pos­si­bil­i­ties are end­less when it comes to the use of data and build­ing mar­ket posi­tions rel­e­vant under anti-trust law.

Data pools and sim­i­lar struc­tures can also facil­i­tate the exchange of infor­ma­tion between com­peti­tors. While this may in some cas­es be pro-com­pet­i­tive, there is, on the oth­er hand, also a risk that agree­ments of this nature could lead to anti-com­pet­i­tive behav­iour, such as price fix­ing or mar­ket shar­ing. Com­pa­nies that hold a dom­i­nant mar­ket posi­tion could abuse this posi­tion by restrict­ing access to their data or set­ting dis­crim­i­na­to­ry con­di­tions for data access. Such behav­iour may be con­sid­ered an abuse of a dom­i­nant mar­ket posi­tion under anti-trust laws.

My opin­ion is:

All this means that the antitrust rel­e­vance of data pools, data lakes and data ware­hous­es lies in their abil­i­ty to cre­ate mar­ket pow­er and influ­ence competition”.

Dr Grace Nacimiento

But – of course, there is a pos­i­tive side which should not be dis­missed- data pools and sim­i­lar struc­tures can also con­tribute to pro­mot­ing inno­va­tion and increas­ing effi­cien­cy. Data shar­ing and access can help com­pa­nies devel­op new prod­ucts and ser­vices and opti­mise their processes.

With the pro­ceed­ings against Face­book, the Ger­man Fed­er­al Car­tel Office has set the first mile­stones in the devel­op­ment of antitrust law with regard to the con­trol of inter­na­tion­al dig­i­tal com­pa­nies with mar­ket pow­er. What was the sub­ject of these pro­ceed­ings and what was the outcome?

The Fed­er­al Car­tel Office con­clud­ed the Meta/Facebook pro­ceed­ings just a few weeks ago with a deci­sion pub­lished on Octo­ber 10. The pro­ceed­ing goes back to 2019 when the Fed­er­al Car­tel Office pro­hib­it­ed Meta (pre­vi­ous­ly Face­book) to com­bine per­son­al user data orig­i­nat­ing from dif­fer­ent sources with­out obtain­ing user  con­sent. Meta filed an appeal against this deci­sion and the pro­ceed­ing went on to the Fed­er­al Court of Jus­tice and the Court of Jus­tice of the Euro­pean Union. Both courts con­firmed the Fed­er­al Car­tel Office’s deci­sion. Pri­or to con­clud­ing the pro­ceed­ings, the Fed­er­al Car­tel Office  and Meta dis­cussed mea­sures to be tak­en for the imple­men­ta­tion of the deci­sion. The mea­sures led to sig­nif­i­cant changes in Meta’s han­dling of per­son­al data of their users. Pri­or to the deci­sion, Face­book could only be used under the con­di­tion that Face­book was allowed to col­lect and com­bine per­son­al data beyond the Face­book ser­vice with per­son­al data gen­er­at­ed in oth­er ser­vices run by Meta, and data gen­er­at­ed in third-par­ty apps and web­sites. As a result of the pro­ceed­ings, users choose which Meta ser­vices they want to link to one anoth­er; they also decide if and which third-par­ty data may be linked to their Face­book data, and, users are now clear­ly informed about their options regard­ing col­lec­tion and use of their per­son­al data by Meta. Specif­i­cal­ly, the mea­sures to be imple­ment­ed by Meta includ­ed in par­tic­u­lar obtain­ing user con­sent  for  Meta to col­lect data from dif­fer­ent Meta ser­vices, estab­lish­ing an Accounts Cen­ter to ensure that data col­lect­ed from dif­fer­ent Meta ser­vices are kept sep­a­rate, intro­duc­ing cook­ie set­tings allow­ing users to decide whether they want their Face­book data com­bined with data which Meta col­lects from third-par­ty web­sites or apps, and ensur­ing con­cise cus­tomer infor­ma­tion to ensure that users are able to find set­tings to avoid data com­bi­na­tion eas­i­ly. All in all, this deci­sion is cer­tain­ly a mile­stone in lim­it­ing the amount and qual­i­ty of data col­lect­ed by large dig­i­tal enter­pris­es and at the same time pro­tect­ing per­son­al user data.

Antitrust laws have already been mod­ernised to a cer­tain extent. Such changes include the 9^th amend­ment to the GWB in 2017, the 2018 coali­tion agree­ment, the estab­lish­ment of a “Com­pe­ti­tion Law 4.0” com­mis­sion with the 10^th amend­ment to the GWB (2021), the so-called “GWB Dig­i­tal­i­sa­tion Act”. What inno­va­tions will they bring? Will they solve press­ing issues and gaps?

The revi­sions or – as you right­ly put it – the mod­erni­sa­tion of the Act against Com­pe­ti­tion Restric­tions was meant to address grow­ing con­cerns relat­ed to the rapid devel­op­ment of Big Data, mean­ing the lit­er­al­ly lim­it­less data col­lec­tion and pro­cess­ing across mar­kets with large cor­po­ra­tions tak­ing a lead in shap­ing busi­ness­es and secur­ing com­pet­i­tive advan­tages based on data use. So, the mod­erni­sa­tion of com­pe­ti­tion law result­ed in a num­ber of impor­tant changes aimed at meet­ing the chal­lenges of the dig­i­tal econ­o­my and strength­en­ing com­pe­ti­tion. Some high­lights of the inno­va­tions include:

The 9^th amend­ment to the GWB in 2017 recog­nised data and access to data as com­pe­ti­tion-rel­e­vant fac­tors in a first sig­nif­i­cant leg­isla­tive step. The 10^th amend­ment to the GWB estab­lished spe­cif­ic data access rights and intro­duced sanc­tions in case of access denials. Com­pe­ti­tion leg­is­la­tion is based on the con­cept of mar­ket pow­er and the appli­ca­tion of the rules requires a pre­vi­ous deter­mi­na­tion of mar­ket pow­er or mar­ket dom­i­nance. Adding to this con­cept of mar­ket pow­er and mar­ket dom­i­nance, the 10^th amend­ment intro­duced the con­cept of “para­mount sig­nif­i­cance for com­pe­ti­tion across mar­kets” to address the grow­ing vis­i­ble pres­ence of inter­na­tion­al dig­i­tal com­pa­nies doing busi­ness across mar­kets. The 10^th amend­ment to the GWB there­by estab­lished a new reg­u­la­to­ry instru­ment for the Fed­er­al Car­tel Office. It enti­tles the Fed­er­al Car­tel Office to inter­vene at an ear­ly stage in order to pro­hib­it anti-com­pet­i­tive prac­tices by com­pa­nies with “para­mount sig­nif­i­cance for com­pe­ti­tion across mar­kets”. Such anti-com­pet­i­tive prac­tices include, for exam­ple, the (ab)use of data pow­er to raise mar­ket entry bar­ri­ers via pro­cess­ing of com­pe­ti­tion-rel­e­vant data col­lect­ed by the respec­tive com­pa­ny and deny­ing or restrict­ing access to such data. The Fed­er­al Car­tel is enti­tled to impose data access oblig­a­tions on the com­pa­nies iden­ti­fied. The new pro­vi­sion also shifts the bur­den of proof to the affect­ed com­pa­nies when it comes to the ques­tion of whether or not there is an objec­tive jus­ti­fi­ca­tion for a cer­tain prac­tice to be pro­hib­it­ed by the Fed­er­al Car­tel Office. It is the com­pa­ny under inves­ti­ga­tion that has to demon­strate that their prac­tice is jus­ti­fied. Such rever­sal of the bur­den of proof is new to tra­di­tion­al abuse con­trol and may well increase the effi­cien­cy of the Fed­er­al Car­tel Office’s pro­ceed­ings. In addi­tion, appeals against respec­tive deci­sions of the Fed­er­al Car­tel Office will now be dealt with direct­ly by the Fed­er­al Court of Jus­tice – instead of going through the High­er Courts first – which may lead to the time­line of legal pro­ceed­ings to reach final court deci­sions being reduced sig­nif­i­cant­ly.    The effec­tive­ness of the new pro­vi­sions and instru­ments has yet to be seen in the long run. They cer­tain­ly do address many of the press­ing issues and gaps in antitrust law, par­tic­u­lar­ly with regard to the chal­lenges of the dig­i­tal econ­o­my. It will be a cru­cial task and chal­lenge for the Fed­er­al Car­tel Office to con­tin­ue review­ing and adapt­ing these rules to meet rapid­ly chang­ing mar­ket con­di­tions. Nev­er­the­less, some suc­cess  can already be chalked up to  this new regime . The Fed­er­al Car­tel Office start­ed inves­ti­ga­tions against inter­na­tion­al dig­i­tal com­pa­nies back in 2021. The first step con­sists in estab­lish­ing the “para­mount sig­nif­i­cance for com­pe­ti­tion across mar­kets” of tar­get com­pa­nies. The Fed­er­al Car­tel Office con­clud­ed this first pro­ce­dur­al step fol­low­ing inves­ti­ga­tions into the mar­ket posi­tions of Microsoft, Alpha­bet, Ama­zon, Apple and Meta and estab­lished  in all cas­es the “para­mount sig­nif­i­cance for com­pe­ti­tion across mar­kets” of these com­pa­nies. On this basis, the Fed­er­al Car­tel Office also con­clud­ed inves­ti­ga­tions of mar­ket prac­tices con­sid­ered anti-com­pet­i­tive. The out­comes led, among oth­ers, to improved pos­si­bil­i­ties for end users to con­trol their data when using, for exam­ple, Google services.

Your chap­ter focus­es on the han­dling of and access to data under antitrust laws. Does this refer to the raw data or to fur­ther processed data deriv­a­tives (infor­ma­tion and knowl­edge) as well? Or is the han­dling of/access to algo­rithms trained with data, neur­al net­works and oth­ers, also rel­e­vant in the con­text of antitrust laws?

Antitrust laws use the term “data”, but do not pro­vide for a statu­to­ry def­i­n­i­tion. So, in prin­ci­ple, the term is to be under­stood in the broad­est sense as includ­ing any type of infor­ma­tion. This means that raw data may be rel­e­vant in the con­text of antitrust laws, but also processed data deriv­a­tives. Algo­rithms and mod­els that have been trained on the basis of large amounts of data can also be rel­e­vant under antitrust laws. Algo­rithms are often the result of exten­sive data pro­cess­ing and can there­fore be a sig­nif­i­cant com­pet­i­tive advan­tage for the data own­er. Sim­i­lar to algo­rithms, neur­al net­works trained on large data sets can also pro­vide a sig­nif­i­cant com­pet­i­tive advan­tage. Access to these mod­els can be rel­e­vant under com­pe­ti­tion law, espe­cial­ly if they are con­sid­ered essen­tial resources for com­pe­ti­tion. From an antitrust per­spec­tive the deci­sive char­ac­ter­is­tic of “data” that is sig­nif­i­cant is their com­pet­i­tive relevance.

Am I right in assum­ing that all kinds of data – whether per­son­al or non-per­son­al – are relevant?

Yes, that is cor­rect. Of course, per­son­al data is the focus of pub­lic aware­ness and atten­tion espe­cial­ly giv­en that end users have become increas­ing­ly sen­si­tive to the use of their per­son­al data by dig­i­tal com­pa­nies. Com­pa­nies must ensure that they com­ply with the legal require­ments for the pro­tec­tion of per­son­al data. Data pro­tec­tion law also may have a role in the deter­mi­na­tion of anti-com­pet­i­tive behav­iour in the rel­e­vant markets.

But, as men­tioned before, “data” refers to any type of infor­ma­tion, so non-per­son­al is rel­e­vant as well. Non-per­son­al data can be used to opti­mise busi­ness process­es, for mar­ket research and to devel­op new prod­ucts and ser­vices. This data is often the basis for analy­sis and research that can lead to inno­va­tion. In addi­tion, com­pa­nies can increase their effi­cien­cy and reduce costs by using non-per­son­al data.

For this rea­son, both types of data are of cru­cial impor­tance for com­pa­nies and com­pe­ti­tion and there­fore are rel­e­vant under com­pe­ti­tion law.

The deci­sive fac­tor is the antitrust aspect of the com­pet­i­tive rel­e­vance of data. When does this cri­te­ri­on apply?

This has to be con­sid­ered on a case-by-case analy­sis. Whether or not data has com­pet­i­tive rel­e­vance depends on a vari­ety of aspects to be tak­en into con­sid­er­a­tion. In the­o­ry, data can be con­sid­ered rel­e­vant if it has any bear­ing on the company’s posi­tion on the mar­ket in ques­tion. Inter­nal and/or exter­nal use of data may influ­ence the goods or ser­vices by improv­ing their qual­i­ty, to name an exam­ple. Data use may also enable the devel­op­ment of prod­ucts in the first place. Oth­er fac­tors to be con­sid­ered when deter­min­ing com­pe­ti­tion rel­e­vance are vol­ume and diver­si­ty of the data, their mar­ket val­ue, their eco­nom­ic val­ue for the com­pa­ny and their  use for devel­op­ing new ser­vices via com­bi­na­tions of data. Also, a fac­tor to be con­sid­ered would be the abil­i­ty of oth­er com­pa­nies to gen­er­ate com­pa­ra­ble data on their own or their pos­si­bil­i­ty to access such data for their own use.

All in all, one can say that com­pet­i­tive rel­e­vance of data depends on var­i­ous fac­tors, includ­ing mar­ket pow­er, net­work effects, inno­va­tion poten­tial, mar­ket struc­ture and reg­u­la­to­ry con­di­tions. These fac­tors will in gen­er­al deter­mine whether and how data affects com­pe­ti­tion in a par­tic­u­lar market.

How is the mar­ket and com­pe­ti­tion deter­mined? Data can be used in many ways and can be copied end­less­ly. A com­mer­cial enter­prise could use aggre­gat­ed GPS data to locate the near­est store, while anoth­er com­pa­ny could use the same GPS data to improve road plan­ning. Both pur­pos­es require the inte­gra­tion of the same or at least sim­i­lar raw data, but their use leads to diverse analy­ses and fur­ther pro­cess­ing which again leads to dif­fer­ent infor­ma­tion and knowl­edge. In view of this, how are the mar­ket and com­peti­tors deter­mined? Is the focus on the indus­try? Or rather on raw data / processed data deriv­a­tives? Or on the pur­pose for which the data is used, based on its legit­imi­sa­tion in the data col­lec­tion? Or on the prod­ucts and ser­vices that are improved or offered based on the data and data derivatives?

Defin­ing the rel­e­vant mar­ket and com­peti­tors in the con­text of data access is com­plex and depends on sev­er­al factors.

The indus­try and the spe­cif­ic use of the data play a cen­tral role in any mar­ket def­i­n­i­tion. For exam­ple, as you men­tioned, GPS data could be used dif­fer­ent­ly in dif­fer­ent indus­tries: in retail to deter­mine the loca­tion of shops and in traf­fic man­age­ment to plan trans­porta­tion infra­struc­ture. A mar­ket def­i­n­i­tion must there­fore, at least in this type of sit­u­a­tion, take into account the spe­cif­ic use of the data.

The type of data may also be a rel­e­vant fac­tor. For exam­ple, both raw data and fur­ther processed data deriv­a­tives may be rel­e­vant. Raw data pro­vides the basis on which var­i­ous analy­ses and appli­ca­tions are devel­oped. Data deriv­a­tives, such as aggre­gat­ed or analysed data, can cre­ate spe­cif­ic mar­kets and com­pet­i­tive sit­u­a­tions. The mar­ket def­i­n­i­tion can in this case there­fore take into account both lev­els, depend­ing on which data is essen­tial for competition.

Then, of course, the prod­ucts and ser­vices that are devel­oped as a result from the use of data are also a cru­cial deter­min­ing fac­tor. If data is used to improve or devel­op new prod­ucts and ser­vices, this may influ­ence how the mar­ket is defined. For exam­ple, a mar­ket for per­son­alised adver­tis­ing based on user data could be defined.

Fur­ther­more, the eco­nom­ic and tech­ni­cal char­ac­ter­is­tics of the data, such as, for exam­ple, its being non-rival and allow­ing mul­ti­ple uses, may have to be tak­en into account. These char­ac­ter­is­tics may also influ­ence the way mar­kets are defined and com­peti­tors identified.

For these rea­sons, the def­i­n­i­tion of mar­kets that are data-based or data-dri­ven requires a com­pre­hen­sive analy­sis of – not only – the fac­tors just mentioned.

Let’s assume that a com­pa­ny has to grant access to its data. Do the data sub­jects then have to waive their infor­ma­tion­al self-deter­mi­na­tion because they can no longer decide for them­selves to whom they want to pass on the data and for what pur­pose? How will the con­flict between antitrust law and data pro­tec­tion be resolved?

The data sub­jects’ infor­ma­tion­al self-deter­mi­na­tion, that is, their right to deter­mine how their per­son­al data is used remains a fun­da­men­tal prin­ci­ple, even in the con­text of antitrust mea­sures. The Gen­er­al Data Pro­tec­tion Reg­u­la­tion (GDPR) ensures, among oth­er things, that per­son­al data may only be processed with the con­sent of the data sub­jects or on anoth­er legal basis. Should a com­pa­ny be oblig­ed to grant access to per­son­al data, this could only be done in accor­dance with data pro­tec­tion reg­u­la­tions. This means that data sub­jects have to be able to con­tin­ue to exer­cise their rights under the GDPR, such as the right to access, rec­ti­fy and erase their data.

When enforc­ing data access rights, the Fed­er­al Car­tel Office and oth­er com­pe­ti­tion author­i­ties must take the data pro­tec­tion rights of data sub­jects into account. This became clear in the pre­vi­ous­ly men­tioned Face­book case where the Fed­er­al Car­tel Office pro­hib­it­ed the link­ing of user data from var­i­ous sources with­out user con­sent. The ECJ con­firmed that antitrust author­i­ties may also take data pro­tec­tion vio­la­tions into account when inves­ti­gat­ing mar­ket abuse in order to deter­mine the abuse of a dom­i­nant mar­ket posi­tion. In doing so, the antitrust author­i­ties must coop­er­ate close­ly with the com­pe­tent data pro­tec­tion author­i­ties and may not exceed their areas of com­pe­tence. So, accord­ing to the ECJ, the antitrust author­i­ties do not take over the tasks of the data pro­tec­tion author­i­ties, but apply the data pro­tec­tion reg­u­la­tions as part of their own inves­ti­ga­tions. At the same time, the ECJ sets clear lim­its with respect to the review of data pro­tec­tion law by antitrust author­i­ties; they may only be car­ried out in con­nec­tion with the deter­mi­na­tion of mar­ket abuse and not as an inde­pen­dent exam­i­na­tion of data pro­tec­tion regulations

For com­pa­nies that have to grant access to data this means n prac­tice must ensure that data pro­cess­ing is car­ried out in accor­dance with the GDPR. This can be achieved through tech­ni­cal and organ­i­sa­tion­al mea­sures that ensure data pro­tec­tion, such as anonymi­sa­tion or pseu­do­nymi­sa­tion of data. The right of data sub­jects to infor­ma­tion­al self-deter­mi­na­tion there­fore remains protected.

Who gets access to the data?

Access to data, par­tic­u­lar­ly in the con­text of antitrust actions, depends on a num­ber of fac­tors, includ­ing the type of data, the legal frame­work and the spe­cif­ic require­ments of the case. Very gen­er­al­ly speak­ing, if a com­pa­ny is oblig­ed to grant access to data asso­ci­at­ed with its dom­i­nant mar­ket posi­tion, it will be the com­peti­tors who are grant­ed access to this data. This is to ensure that com­pe­ti­tion in the mar­ket is not hin­dered by exclu­sive access to impor­tant data. In some cas­es, third par­ties, for exam­ple, research insti­tu­tions or part­ner com­pa­nies, may also be grant­ed access to the data if and to the extent this is nec­es­sary to pro­mote inno­va­tion or to com­ply with legal requirements.

How does access work? The release of a copy of the data? A right of access to the data pool or part of the data pool? Chang­ing the data pool from a “pro­pri­etary” to an “open” regime…?

Access to data in the con­text of antitrust mea­sures can be imple­ment­ed in dif­fer­ent ways, depend­ing on the spe­cif­ic cir­cum­stances of the case and the require­ments of the com­pe­ti­tion authorities.

In some cas­es, for exam­ple, it may be nec­es­sary for the com­pa­ny which is faced with this sort of oblig­a­tion to hand over a copy of the rel­e­vant data to the com­peti­tor. This allows the com­peti­tor to analyse and use the data inde­pen­dent­ly. A dif­fer­ent sce­nario could be a right of access to a data pool or a part of a data pool. This means that a com­peti­tor, or oth­er third par­ty, obtains direct access to the data­base and is able to retrieve the data in real time or at peri­od­ic inter­vals. This may be par­tic­u­lar­ly rel­e­vant if the data needs to be con­tin­u­ous­ly updat­ed. In oth­er cas­es, it may in fact be nec­es­sary to change the data pool from a pro­pri­etary to an open regime. This would mean that the data is made pub­licly avail­able so that not only spe­cif­ic com­peti­tors but also oth­er mar­ket par­tic­i­pants and the pub­lic have access to the data. How­ev­er, this can be com­plex and would require in par­tic­u­lar care­ful con­sid­er­a­tion of data pro­tec­tion and trade secrets.

Regard­less of the type of access that has to be grant­ed, it is impor­tant to remem­ber that tech­ni­cal and organ­i­sa­tion­al mea­sures must be put in place to ensure data pro­tec­tion. This may include anonymi­sa­tion or pseu­do­nymi­sa­tion of data to ensure that per­son­al data remains protected.

Does the com­pa­ny’s access to its data make it a gate­keep­er accord­ing to the def­i­n­i­tion found in the Dig­i­tal Mar­kets Act?

The Dig­i­tal Mar­kets Act defines gate­keep­ers as large dig­i­tal plat­forms that play a sig­nif­i­cant role in the sin­gle mar­ket, pro­vide an impor­tant inter­face for busi­ness users and hold, or are like­ly to hold, an estab­lished and sus­tain­able mar­ket posi­tion. Exam­ples of such gate­keep­ers are com­pa­nies such as Alpha­bet, Ama­zon, Apple, ByteDance, Meta and Microsoft.

A com­pa­ny that is oblig­ed to grant access to its data does not auto­mat­i­cal­ly qual­i­fy as a gate­keep­er. The gate­keep­er sta­tus rather depends on cer­tain cri­te­ria and not sole­ly on whether the com­pa­ny grants access to data. In order to be des­ig­nat­ed as gate­keep­er, sev­er­al cri­te­ria must be met. Very gen­er­al­ly speak­ing, the company’s plat­form must have a sig­nif­i­cant impact on the inter­nal mar­ket. The plat­form must be an impor­tant inter­face for busi­ness users to reach end users, and it must have or be like­ly to achieve an estab­lished and sus­tain­able mar­ket position.

What jus­ti­fi­ca­tion could a com­pa­ny use to counter the right of oth­ers to access their data? Prop­er­ty rights to “its” data? Pro­tec­tion of its invest­ments in dig­i­tal inno­va­tion? Safe­guard­ing the rights of data sub­jects with regard to per­son­al data?

A com­pa­ny could use var­i­ous jus­ti­fi­ca­tions to oppose a data access right.

Com­pa­nies could, for exam­ple, argue that they have pro­pri­etary rights to the data they col­lect and process. They could argue that the data is a valu­able busi­ness asset which was obtained through sig­nif­i­cant invest­ment in tech­nol­o­gy and infra­struc­ture. Anoth­er argu­ment could be the pro­tec­tion of invest­ments in dig­i­tal inno­va­tion. Com­pa­nies often invest sig­nif­i­cant resources in the devel­op­ment of data col­lec­tion and analy­sis tech­nolo­gies. So they may argue that imposed access to this data could reduce the incen­tives for such invest­ments and weak­en the com­pa­ny’s inno­v­a­tive capa­bil­i­ty. The pro­tec­tion of data sub­jects’ rights under the Gen­er­al Data Pro­tec­tion Reg­u­la­tion (GDPR) could be anoth­er strong argu­ment. Com­pa­nies could claim that the dis­clo­sure of per­son­al data with­out the explic­it con­sent of the data sub­jects vio­lates data pro­tec­tion reg­u­la­tions and jeop­ar­dis­es the infor­mal self-deter­mi­na­tion of the data sub­ject. This could be a rel­e­vant argu­ment if the data can­not be anonymised or pseu­do­nymised. Also, com­pa­nies could point to tech­ni­cal and organ­i­sa­tion­al chal­lenges relat­ed to hav­ing to pro­vide access to data. This could include the need to adapt com­plex IT infra­struc­tures or imple­ment addi­tion­al secu­ri­ty mea­sures to ensure data protection.

In any case, these and any oth­er argu­ments against grant­i­ng access must be care­ful­ly weighed against the objec­tives of com­pe­ti­tion law to ensure that com­pe­ti­tion is pro­mot­ed with­out infring­ing the rights and inter­ests of the com­pa­nies and indi­vid­u­als concerned.

In his Duet inter­view, Hans Jür­gen Jakobs clear­ly describes the com­pe­ti­tion and the bat­tle for monop­o­lies. Can Europe still hold its own in the cur­rent data and dig­i­tal­i­sa­tion mar­ket? Some ways for estab­lish­ing and main­tain­ing monop­o­lies appear to be sub­si­dies, the pro­mo­tion of  tech­nol­o­gy and price wars. Must or should Europe enter this com­pe­ti­tion or should Europe fight with oth­er means such as antitrust mea­sures, import and export duties among oth­er ways in order to be or become a seri­ous com­peti­tor? Europe’s leg­isla­tive bod­ies in par­tic­u­lar are  seem­ing­ly “lead­ing” the new chal­lenges in terms of data pro­tec­tion, AI, etc., mean­ing they are “show­ing” the way. Is this Europe’s only “weapon” against dig­i­tal monop­o­lies? If so, how effec­tive is this “weapon”?

Europe is in fact fac­ing sig­nif­i­cant chal­lenges with regard to glob­al com­pe­ti­tion in terms of dom­i­nance in the dig­i­tal mar­ket. There are var­i­ous means for Europe to meet the chal­lenges and assert itself.

First of all, Europe is already invest­ing heav­i­ly in dig­i­tal­i­sa­tion and tech­no­log­i­cal inno­va­tion. Pro­grams such as the Dig­i­tal Europe Pro­gramme are aimed at accel­er­at­ing dig­i­tal trans­for­ma­tion and strength­en­ing dig­i­tal capac­i­ty in key areas such as arti­fi­cial intel­li­gence (AI), cyber­se­cu­ri­ty and high-per­for­mance com­put­ing. The Pro­gramme pro­vides fund­ing in areas which are cru­cial to Europe’s posi­tion, includ­ing, for exam­ple, arti­fi­cial intel­li­gence and cyber­se­cu­ri­ty. These invest­ments are vital for boost­ing the com­pet­i­tive­ness of Euro­pean com­pa­nies and dri­ve innovation.

More­over, antitrust mea­sures are also a key fac­tor in this con­text. With the Dig­i­tal Mar­kets Act and the Dig­i­tal Ser­vices Act, Europe has cre­at­ed strong reg­u­la­to­ry instru­ments to pro­mote com­pe­ti­tion and lim­it the pow­er of large dig­i­tal plat­forms. These laws aim at cre­at­ing a lev­el play­ing field and pre­vent abuse by dom­i­nant com­pa­nies. The enforce­ment of these reg­u­la­tions is an impor­tant means at com­bat­ing dig­i­tal monopolies.

Import and export duties could in the­o­ry also be used as a means of influ­enc­ing com­pe­ti­tion. How­ev­er, such mea­sures are often high­ly con­tro­ver­sial and can cre­ate trade con­flicts. If at all, they should there­fore be used with cau­tion and in accor­dance with inter­na­tion­al trade agreements.

As we know, Europe has estab­lished itself as a pio­neer in the area of data pro­tec­tion and eth­i­cal stan­dards. The GDPR sets glob­al bench­marks and forces com­pa­nies to com­ply with high data pro­tec­tion stan­dards. This reg­u­la­to­ry frame­work can be used as a sig­nif­i­cant com­pet­i­tive advan­tage by cre­at­ing trust among con­sumers and pro­mot­ing Euro­pean val­ues in the dig­i­tal econ­o­my. Fur­ther­more, Europe can strength­en its posi­tion through inter­na­tion­al coop­er­a­tion. This includes part­ner­ships with oth­er coun­tries and regions to devel­op com­mon stan­dards and pro­mote glob­al trade. In the end, there is a com­bi­na­tion of dif­fer­ent means and tools for Europe to com­pete inter­na­tion­al­ly in the dig­i­tal econ­o­my. Cre­at­ing a lev­el play­ing field and encour­ag­ing inno­va­tion is key to assert­ing a lead­ing role. At the same time, it is cru­cial to con­tin­u­ous­ly adapt and evolve these poli­cies to meet the rapid­ly chang­ing chal­lenges of the dig­i­tal economy.

Grace, thank you for shar­ing your insights on com­pe­ti­tion aspects of data warehouses.

Thank you, Cristi­na, and I look for­ward to read­ing your upcom­ing inter­views with recog­nised experts, delv­ing even deep­er into this fas­ci­nat­ing topic.