Yesterday @EU_Commission published the proposal for Data Governance Act #DGA. Let's critically unpack the file. The set of measures is a good step forward, but we are still not discussing public value and competition enough. Thread 1/12 #data #digital #GDPR

A @Deloitte impact assesment study claims that the data economy will grow to EUR 544.4 bln EUR (3.95% GDP). Previous studies by that very same company for GDPR mentioned decreases of a similar magnitude, which never happened. Instead of numbers let's talk who gets value. 2/12

First of all, #DGA lacks symmetry. It clearly aims for 'making public data available to re-use', but for private data it advocates only 'sharing between businesses'. Is data altruism enough to tap into the huge datasets held by private companies? Very risky to believe so. 3/12

Art 4 allows exclusivity in rights to re-use of public data for 3 y if neccessary. But why would it ever be "neccessary"? Seems like a way for Big Tech to claim miracles and restrict new entrants. Wasn't DeepMind and NHS enough of a lesson? Or am I getting it wrong? 4/12

There is still a heated debate on #datalocalization and #WTO unrestricted transfers. But WTO has no legislative power. The so-called "rules" are actually e-commerce trade deals. Last year the G20 Osaka track was blown up for data colonialism. WTO will face similar scrutiny. 5/12

Conditions for re-use should go from "may" to "should" impose obligations on re-use of data in a secure environment provided by public sector. MIT OPAL and data commons studies show that moving algorithm to data is the most secure option. 6/12

This will end the #datalocalization debate as well. Since the data stay in the EU, and the algorithm is sent, no data transfer happens. Algorithm IPRs will be secure if public data commons are trusted. But homomorphic encryption & federated learning also come in handy. 7/12

Let's discuss competition. Data re-use by Big Tech DECREASES comp. (see @PAghion). So "non-discriminatory" and "shall not restrict competition" are contradictory! But "proportionate" & "justified" prove that fees can be progressive - the bigger you are, the more you pay. 8/12

Great that sharing providers cannot sell data. But asking for non-discrimination "including as regards prices" is absurd. Big Tech has to pay more: it has more capacity and decreases competition. Progressive data fees are key to prevent value leakage from public/coop sector. 9/12

Instead of a new regulator, EC plans to create an expert group (European Data Innovation Board) to facilitate best practices. Hope that it will bring the best of European diversity! Young people deserve representation as well. We were born & raised in data economy 10/12

Common data altruism consent form is great. Lowering the barriers of entry. However, the "form" should come together with open exchange protocol, a new standard for data portability. It should be mandatory for data exchange. Data altruism & enforcing GDPR with one stone 11/12

And finally & once again: the #DGA lacks B2G data sharing. Where is mandatory sharing for Big Tech? What about municipal, healthcare, all other datasets of public interest? It is crucial our, European peoples', data belongs to us and is commonly stewarded by us. Period. 12/12

CC: @W_Wiewiorowski @ZagorskiMarek @JuRomanowska @PaulNemitz @ThierryBreton @vmanancourt @miapetrakumpula @Melissahei @CjColclough @maxschrems @airam_ainafets @KartaCyfrowa @laura_zoboli