Dear readers,
The summer has reached Brussels – in theory, but not always in practice. As we gear up to leave the rain behind in favour of warmer corners of the continent, we are also approaching a new EU legislative term. Below is how we expect tech policy to look in this new era.
The focus on implementation
First of all, the general approach will change. After a quite substantial amount of digital legislation in the current term, with the adoption of the Data Act, Data Governance Act, Digital Services Act, Digital Markets Act, Cyber Resilience Act, Chips Act, NIS2 and of course the AI Act, the focus will now move on to implementation. Since many of these files overlap, one challenge of the process will be to avoid inconsistencies in how all these sets of rules are applied.
The focus on implementation doesn’t mean we are going to get bored, quite the contrary. Starting from the recently-established AI Office, the AI Pact will be in charge of helping companies to start complying with the AI Act measures in the next 2 years before it takes effect.
In terms of secondary legislation, we are expecting around 20 delegated and implementing acts that will go into the details of the AI Act. The European Commission is also set to publish guidelines on the implementation of the file.
Another key aspect is standardisation, carried out by CEN and CENELEC, which will have a massive impact on how the AI Act will be applied in practice. The AI Liability Directive will complete the framework. It was put on hold in the previous term, but is likely to resume soon.
Other new initiatives on the radar
But it’s not just implementation – some new proposals are waiting in the pipeline.
The Digital Networks Act is the one new big initiative expected for the next term. It will address the concerns of the telecom sector of the past years such as market concentration, spectrum management and investments in networks. It will also be a key pillar of a new competitiveness strategy as a reinforced digital single market will mean a more digitalised – and therefore competitive – digital industry.
While the Commission and strategic documents like the Letta report seem to go into that direction, a resistance at the Member States’ level, reinforced by the results of the elections, may slow down this process and push in the opposite direction.
The GDPR Review is expected soon. While the first report considered the regulation fit for purpose, several issues related to its application have emerged in the past years. In addition, the dramatically changing legislative framework makes the review more likely in the new term. Whether it will be a complete investigation (aka opening Pandora’s box) or a set of targeted amendments, like the ongoing proposal on GDPR enforcement in cross-border cases, we will find out in the next months.
Last but not least – keep an eye out for virtual worlds. The institutions have shown great interest in regulating the metaverse and considering its potential benefits. The Commission published a strategy on virtual worlds in 2023. The Parliament adopted two own-initiative reports dealing with consumer/data protection and civil/commercial law aspects of the metaverses. While a legislative initiative has not been announced yet, it is one of the few things that haven’t been regulated and – in light of the results of the elections – one of the areas that are likely to be supported across the different political groups.
Enjoy your break – we’ll see you in September – and before that, check out the latest developments in tech we prepared for you below.
Best,
Giovanni
Love reading out-of-the-Brussels-bubble tech news? Share it with your friends and colleagues. They can subscribe here. Ideas? Suggestions? Comments? Send them our way!
#TechAways is brought to you by SEC Newgate EU’s one and only team featuring Julia Piwowarska, Camilla Frison, Alice Palumbo, Giovanni Bazzoli and Ali El Majjaoui.
More space for data way above the clouds 💾 🌥I [The Next Web]
With AI everywhere, the demand for more data storage is growing rapidly; and this is just the beginning. The EU-funded ASCEND study came up with an unexpected place to build data centres: space. The aim is to strengthen Europe’s data sovereignty, achieve net-zero and address the environmental impact of AI’s massive energy consumption. While it sounds like the plot of Wall-E (2008), where futuristic space technologies solve the Earth’s environmental degradation, it actually makes a lot of sense. First of all, there is no sunnier place than space, providing these data centres with 24/7 solar energy. The freezing space temperatures also mean there is no need for a large quantity of water in the cooling system. Though, not everything is as green as it seems. The environmental impact of launching a rocket is still significant. Therefore, it will be crucial to design a launch technology with 10x fewer emissions and develop a rocket fuel that keep these data centres in orbit. Several companies are already working on developing these eco-friendly alternatives to be implemented by 2036.
The Iron Leg: MIT’s brain-controlled prosthesis breakthrough 🦿 [Yahoo! News]
In a breakthrough reminiscent of the Stark Industries project, MIT researchers have developed the first leg prosthesis fully controlled by the brain, achieving a natural stride. The study, involving seven patients, used “agonist antagonist myoneural interface” (AMI) surgery, which allows users to sense limb position, speed, and torque. Unlike conventional prosthetics controlled by robotic algorithms, this new design translates brain signals directly into movement. This innovation enables users to walk naturally on varied surfaces, including slopes and stairs. Co-author Hugh Herr, a double amputee, is considering the surgery himself. This breakthrough represents a significant advancement in restoring mobility for those with severe limb injuries and a giant (iron) leap forward!
Artificial undergraduates 🤖📚 [Ars Technica]
Chat GPT 1 – Human brain 0. In an outstanding result for Chat GPT, the software outsmarted college freshmen in intro-level courses. The AI detection software failed miserably in the University of Reading’s large-scale experiment, where fake students submitted AI-generated essays, showing that ChatGPT aced exams better than real students. Still, ChatGPT was unable to keep up in advanced classes. It is still unable to do actual critical thinking or in-depth analysis, which is why final-year students performed better on those challenging tests. With continuous improvement, the software may develop those skills to score better in advanced assignments in the future. Since education’s purpose is to prepare students for their professional careers, AI will soon be integrated into the classrooms. After graduation, students will undoubtedly need to use a variety of AI technologies, and they better know how to use them properly.
In case you haven’t had enough:
- Eye-tracking VR system helps children stay still during MRI scans [The Next Web]
- AI companies are finally being forced to cough up for training data [MIT Technology Review]
- Japan wins 2-year “war on floppy disks,” kills regulations requiring old tech [ArsTechnica]
- AI Is Rewriting Meme History [WIRED]
About this week’s editor, Giovanni Bazzoli:
I lead SEC Newgate EU’s tech policy team. As you can guess from my name, I’m 100% Italian, but have been in town for a few of years now. In my spare time, I enjoy running and studying the topics I deal with every day from an academic perspective as a teaching assistant at an Italian university.