Two years of trying – and failing – to get government consensus

Proposed in 2022 by the EU’s outgoing home affairs chief, Ylva Johansson, the draft EU CSA Regulation sent shockwaves throughout the privacy and human rights communities.

The European Commission tried to downplay the massive threat this law, coined “Chat Control”, would pose to the security of everyone’s digital communications. But soon, they had to reckon with intense criticism from every angle: civil society, legal experts, technologists, child rights specialists, and police, warning that the Regulation would be counter-productive to its stated aim of protecting children.

The EU’s independent legal services also joined in with damning criticism of the draft law, confirming that the law would mean such a massive violation of innocent people’s digital human rights, that the EU’s top court would surely strike it down.

This was an important set of developments because in order to pass this controversial bill, the European Commission would need the support of both the European Parliament and the Council of the EU (the body representing the governments of all 27 EU Member States).

Yet for over two years, the Council of the EU has been unable to find a position that satisfies both the EU countries that (thankfully) appreciate the importance of digital security – and those whose terrifying mission is to break encryption and put an end the presumption of innocence online.

Orbán pushes for backdoors into Europe’s private chats

In summer 2024, the government of Hungary became the fifth country to be given the unenviable task of attempting to broker a common position of the Council of the EU on this ill-fated law. The European Commission has long been trying to convince Member State governments that the proposed Regulation is legally sound (it isn’t), would protect encryption (it wouldn’t) and that reliable technologies already exist (they don’t).

Thankfully, leaked documents from September 2024 reveal that countries including Germany, Poland, Austria, Estonia, Slovenia and Luxembourg remain firm in their insistence that any such law must still comply with human rights and technological reality. Their leaders have been clear that until the proposed Council position meets these criteria, they cannot agree to it.

But pressure has been mounting on the countries that have been on the fence – apparently including the Netherlands, France and Italy – to say “yes” to a proposal that would give police a back door into everyone’s private digital chats.

According to Politico and to local reports, notorious Hungarian Prime Minister, Viktor Orbán, pulled out all the stops to try and convince the Netherlands to support the latest text. And in the last few days, he came worryingly close to succeeding.

Surprise as spooks come to the rescue

In late September, digital rights groups were shocked to hear reports that the Dutch government was suddenly considering endorsing Hungary’s proposal. The Netherlands had previously committed that they would not support any proposal on this law until they could be sure that it would protect encryption. But they may have been swayed by pervasive false claims that the latest proposal, coined “upload moderation”, protects encryption.

Over three hundred researchers and scientists from across Europe and beyond quickly jumped in to (once again) remind all EU governments that this completely contradicts widespread expert consensus. There is no way to systematically scan and report on private messages and keep those messages secure, regardless of whether you do this scanning on a person’s device or elsewhere.

On 1 October, following significant mobilisation from civil society, including EDRi member Bits of Freedom and national opposition politicians, the news broke that the Netherlands would officially abstain from the proposal. This is a welcome development, because it means that Hungary does not have a majority to move forward with their proposal, instead having to remove the CSA Regulation from an upcoming Council agenda.

One of the most interesting parts of the Netherlands’ will-they-won’t-they saga, however, is the fact that one decisive element seems to be an opinion of the national security service. Dutch spooks warned their government that the latest proposal would threaten the cybersecurity of the country, putting national security at risk. This is a warning that should resonate with other countries, too.

Dead but not buried (yet)

It’s hard to see where Hungary can go from here. Despite repeated attempts by several countries, it has once again been shown to be impossible to have a law that protects encryption whilst simultaneously undermining it, nor to legalise illegal mass surveillance.

This should be the final nail in the coffin for this Frankenstein’s Monster of a law, which EDRi has warned since the beginning is legally and technically infeasible. No amount of linguistic gymnastics can change that.

But this file has come back from seemingly being dead several times already in recent years. Since the very advent of encryption, governments have been arguing that they should be able to have widespread access to the encrypted communications of people even without reasonable suspicion.

With a new Commission taking up the charge at the end of 2024, this is an opportunity to press the reset button, and start again. We deserve a better proposal which would tackle the grave issue of child sexual abuse in a way that complies with EU human rights law. One which would ensure a safe internet for all, rather than causing even more harm.

Read the full article here.

Gerelateerd nieuws

Europese waakhond: ‘AI‑systemen bedreigen grondrechten, menselijk toezicht onvoldoende’

De Europese grondrechtenwaakhond FRA (European Union Agency for Fundamental Rights) waarschuwt in een vorige week verschenen rapport dat organisaties slecht zijn voorbereid op het beoordelen en beperken van grondrechtenrisico’s bij het gebruik van hoog‑risico‑AI. Volgens de FRA dreigt daardoor een kloof tussen de ambities van de AI Act en de dagelijkse praktijk bij ontwikkelaars en gebruikers van AI‑systemen in onder meer asiel, onderwijs, werk, politie en sociale zekerheid. Die kloof raakt direct aan de manier waarop mensen en AI in de samenleving samen optrekken: als de menselijke kant van die samenwerking – kennis, reflectie en kritisch vermogen – tekortschiet, verliest AI haar grond voor vertrouwen.

De cruciale rol van een veilige meldcultuur bij cybersecurity

Onlangs publiceerde de Nationaal Coördinator Terrorismebestrijding en Veiligheid (NCTV) het Cybersecuritybeeld Nederland (CSBN) 2025. Het rapport schetst een digitale dreigingsomgeving die steeds complexer, diverser en onvoorspelbaarder wordt. Terwijl de dreiging groeit, ligt de verdediging in het versterken van de digitale basishygiëne. Organisatiecultuur speelt hierbij een belangrijke rol in hoe incidenten worden herkend en gemeld. Het CSBN 2025 maakt duidelijk dat digitale veiligheid geen puur technologisch vraagstuk is, maar afhankelijk is van hoe mensen binnen organisaties handelen. Transparency International Nederland (TI-NL) benadrukt daarom de cruciale rol van menselijk gedrag en een veilige meldcultuur bij effectieve cybersecurity.

AI en Auteursrecht: waarom een uitspraak uit München alles verandert

Op 11 november 2025 deed het Landesgericht München uitspraak in een zaak die de juridische wereld én de techsector op scherp zet: GEMA tegen OpenAI (zaaknummer 42 O 14139/24). Het ging om de vraag of het gebruik van auteursrechtelijk beschermde songteksten door generatieve AI-modellen zoals ChatGPT in strijd is met het auteursrecht. Het antwoord van de rechtbank? Ja. En dat is best baanbrekend.

Topbestuurders zien AI als oplossing voor klimaatambities, niet als bedreiging

Topbestuurders zien Kunstmatige Intelligentie (AI) als belangrijk hulpmiddel voor het behalen van klimaatdoelen. Dat blijkt uit een wereldwijd onderzoek van KPMG onder ruim 1.200 CEO’s en bestuursvoorzitters in twintig landen. De meeste ondervraagde topbestuurders (87 procent) geven aan dat AI-toepassingen bijdragen aan het realiseren van net-zero-doelen. Hoewel slechts 30 procent van de organisaties op korte termijn prioriteit geeft aan het verbeteren van het energieverbruik door AI, is 96 procent ervan overtuigd dat er in de toekomst voldoende duurzame energie beschikbaar zal zijn om aan de energievraag van AI te voldoen.