Main takeaways
- The EU is rushing negotiations on tech and digital files at the expense of quality
- Last-minute deals and radical changes harm EU’s credibility and create uncertainty
- Lawmakers need to stop rushed deal-making and focus on better policy instead
Brussels seems to have been taken hostage by an unsettling trend of last-minute EU deal-making on many major tech and digital files. Even compared to a few years ago, almost every recent file agreed has been rushed at the finish line – often at the expense of quality and dismissing months, or even years, of diligent work. This unfortunate trend is harming the EU’s credibility, a message which so many stakeholders keep voicing behind closed doors in shared frustration.
Particularly in recent months, EU institutions and Council Presidencies have been so desperate to announce a ‘deal’ on a big file that they rushed negotiations at any price. We’ve now seen this happen in Brussels too many times to count or ignore. Provisional political deals were struck, victory declared, and then significant changes were later introduced to the final legal text. This is of course not a new trend. It happened before with the Digital Services Act, but you would expect EU lawmakers to learn from their mistakes.
Yet, in recent months these same mistakes have been repeated again: with the EU’s Artificial Intelligence (AI) Act, the European Media Freedom Act, new political ad rules, the Data Act, and most recently the Platform Work Directive. Negotiators strike a hasty deal presuming they can salvage a workable law in the aftermath, but this part of the process is hardly subject to public scrutiny or proper impact assessments. It can lead to confusion that undermines the entire process and goes against the desired outcome.
Unfulfilled promise to do better
In 2016, the EU institutions actually signed an interinstitutional agreement on better law-making, in which they committed to upholding key principles to improve the legislative process, ranging from more transparency to better impact assessments. I applauded it back then, and dare I say many in industry looked forward to its positive trickle on every sector. For example, it was agreed that “substantial amendments” to the Commission’s original legislative proposal by the co-legislators should be subject to impact assessments, which is very logical.
Unfortunately, we have not seen any such exercise in recent digital policy files, and definitely not the AI Act. Europe works on some of the biggest files that impact all our daily lives, but the biggest changes are sometimes made as the wind blows.
The EU wanted to be ‘the first’ to regulate AI, and repeated that mantra enough to make the priority clear. But no one really focused on being ‘the best’. That wasn’t part of the narrative. If you look at the political agreement struck on the AI Act, you will see that it introduces new rules and concepts that were not properly discussed, nor part of the co-legislators’ negotiating mandates. But these last-minute additions greatly impact how AI will be developed and used in Europe, and possibly the world, for years to come.
Harming the EU’s credibility
It’s worrisome to see that negotiators divert so much from democratically agreed mandates to strike a deal and get media coverage. Is it political pressure from the top? Are they doing this in an attempt to get more votes or show constituents that they are ‘delivering’ for them? Because if that’s the case, voters might conclude the opposite when they learn about this kind of deal-making behind closed doors, void of real scrutiny.
What happened to proper EU trilogue negotiations, celebrated for convergence and consensus between institutions based on agreed mandates? I still have trouble understanding how negotiators can ignore years of work by introducing radically-different concepts, often with open-ended language that is only acceptable because each party interprets it completely differently, at the last minute. How can any of us (including the negotiators themselves) predict the impact of such decisions, if those photo ops need to be followed by months of further work to solve outstanding problems?
Half-baked ‘solutions’
This often results in half-baked ‘solutions’ that create new problems, or envision completely new concepts that don’t correspond with reality or established definitions. It happened with the Data Act last year. When the EU institutions’ self-imposed deadline of six months started to approach, negotiators agreed on a pastiche piece of legislation covering a bit of everything. What we now have is a shaky set of rules that raises far more questions than they answer, be it on trade secret protection, data flows, or government access to business data, to name a few. Now it’s likely up to the EU Court of Justice to answer those questions in a few years. In the meantime, how do you implement properly with such high stakes?
While December might have been overshadowed by the AI Act, it also produced last-minute deals on the revision of the Product Liability Directive and the European Media Freedom Act, which are other examples of how rushed deal-making leads to unintended consequences. Indeed, while the Digital Services Act was supposed to become a tool to harmonise the Digital Single Market, it’s already being challenged – and even contradicted – by other EU laws, including said Media Freedom Act. And the new product liability rules could actually end up disincentivising the very AI innovation that the AI Act claims to promote.
Platform work mess
Even as recently as this month the trend continued. The proposed Platform Work Directive (PWD), which has been in trilogue negotiations for several months now, was the latest victim. The Spanish EU Presidency actually already declared ‘victory’ on the PWD back in December, only for EU Member States to subsequently pull the plug on the Spanish deal.
In an attempt to break the deadlock, the Belgian EU Presidency all of the sudden (unilaterally) decided to remove a key feature from the draft PWD in early February: the so-called ‘presumption criteria’. This actually was, until recently at least, the primary mechanism to fulfil the Directive’s goal of creating legal certainty.
The very idea of removing this key provision – simply because political agreement wasn’t expedient, and with EU elections around the corner – rendered the point of the Directive completely moot, as it basically would end up contradicting the EU Treaties. The half-baked solution that was tabled would literally have left all key issues to be defined by individual EU countries, which basically meant the PWD itself would become pointless.
Ironically, national governments also rejected this last-ditch attempt the other week. Whether the Belgians will manage to salvage the gig work bill remains to be seen. However, simply by tabling a radical ‘compromise’ that ignored almost three years of discussion, any future deal is likely to be further removed than ever from the EU institutions’ democratically agreed mandates. It’s another dangerous example of EU lawmakers rushing to an outcome, regardless of what the outcome does or the harm it could create.
Time to start walking the talk
That’s why in 2024, I hope that stakeholders far and wide will join forces to make clear to newly elected Members of the European Parliament and the next European Commission that this is not the way forward for a positive Europe. If the private voices from all institutions who truly want better can be heard, industry is there with them. Indeed, the EU needs better policy-making to create an effective and coherent legislative framework. Not last-minute politics or more of these self-serving selfie moments.
Better regulation should be a key cornerstone of policy-making, and it’s time the EU institutions really start walking the talk when they know what’s truly needed. Responsibility and accountability work both ways, and industry expects more.
An abridged, German version of this article first appeared in Tagesspiegel Background.