This article is part of our special report Mapping Europe’s tech landscape.
The EU’s digital agenda over the next mandate is set to be marked by a series of broad-ranging reforms, from artificial intelligence and data protection to cryptocurrency regulation and digital tax. EURACTIV talked to Greek MEP Eva Kaili about how she hopes the EU’s digital agenda over the next five years will play out.
Eva Kaili is a Greek MEP for the Socialists and Democrats group. Ahead of her keynote address at the upcoming EIT Digital Conference, she responded to questions from EURACTIV’s Digital Editor Samuel Stolton.
Let’s start with one of the more contentious issues of the next mandate. What’s your take on ethics in the field of Artificial Intelligence? Would you welcome regulation in Artificial Intelligence to help preserve EU principles? What are the challenges of this?
I think the ethical aspect of AI should be a primary concern. Already in the European Parliament report on AI which I shadowed on behalf of the S&D group, we focused a lot on the ethics of AI development. On many paragraphs, we emphasized on “ethics by design” approach and on the fact that companies producing AI must be very cautious on incorporating any kind of bias on the AI systems.
We also warned against any malicious use of AI from companies who might use face recognition or mood detecting AI technologies in order to manipulate your emotions, feelings and try to gain profit from them by selling you relevant products.
Another issue that will be very important is sensitive personal data. We have to ensure that under no circumstances any AI system will be able to disclose any sensitive private data it might collect. The right to privacy will be one of the most pressing issues in the years to come in my opinion and AI will have to be adjustable to this.
How concerned are you with the use of AI across totalitarian regimes worldwide?
Very. But even more worrying is what those regimes will be able to do to their citizens by deploying advanced AI surveillance systems. Will it be then that we will see a pure Orwellian universe where the state will be able to control even the emotions of its people?
Another concern is weaponised AI. Let’s just imagine how much destruction they will be able to cause. My intention is not to set the scene of a bleak future but to affirm that we will need international cooperation and specific standards exactly the same way we have treaties for chemical and biological weapons.
You recently submitted a question to the European Commission in which you highlight potential challenges with regards to GDPR hindering AI development. Could you say a bit more about this?
The GDPR is a landmark legislation which established a framework that safeguards our data from being processed without our consent – giving us back our privacy and control over personal information ownership. Although a great achievement, the GDPR includes provisions which could potentially hinder the development of data-driven, self-learning algorithmic systems.
For example, if Article 22 of the GDPR is applied to AI tools, automated decision-making systems would be faced with regulatory backstops aimed at safeguarding the rights of data subjects.
AI development is a cornerstone of the EU’s future strategy. It is important to create a space for it, in which it can be developed – and its development relies on the availability of datasets. It is therefore important to combine the lessons learned from GDPR in the AI space, without over-regulating it so that innovation can be incubated rather than halted.
G7 ministers recently put forward a series of requirements for Facebook’s new coin ‘Libra’ to abide by. Do you agree that there are ‘serious risks’ surrounding Libra? What could these be and how could these be met at an EU level?
Libra is just one more example of a corporation that tries to leverage the tokenization of value in its business model. Many firms have done it already and thousands will follow soon. This is a natural evolution of the economy and we cannot stop the democratization of money.
I introduced those concepts two years ago in the DLTs and Blockchain Resolution. Unfortunately, ECB, EBA and ESMA ignored my recommendations, they adopted a rather defensive view on crypto assets and they systematically try to keep under the carpet a wave of disruption. Well, the risks are there, but the potential is immense as well.
I believe that we need a holistic view about the tokenization of the economy, to tame the related risk by making a crypto assets integral part of the mainstream regulated economy, and re-assess the idea of monetary policy and central banking by figuring out their relevance in an economy where IoT and AI play a dominant role.
You’ve previously been heavily involved in tax issues. What is your opinion on the taxation of digital giants? Do you agree that there is a risk of the EU market fracturing if a common EU approach isn’t found on digital services tax? Or would you prefer a broader international agreement?
Digital tax is a sine qua non in the current economic reality. The idea behind it, though, should be rather positive than negative. In my view, digital tax is a tool for reduction of the digital inequality within the economy and the creation of competitive advantage between the economies. Also, it is a tool for bridging the digital gap within the society and the business field with critical investments in digital infrastructure, digital innovation and what I call “human digital capital” through education.
Digital Tax should not be a burden but a factor of acceleration of the competitive position of the EU economy as well as a factor of accelerating the transition to Industry 4.0. For this reason, I suggest an EU Digital Transition Fund where proceedings of the Digital Tax will be accumulated and then channelled in digital transition projects.
Allow me to predict that the current proposition will never be adopted. Digital Tax is a necessary growth instrument that requires us to be ambitious, technologically savvy, and competition relevant. A good European Digital Tax will provide EU with a significant regulatory advantage over our global competitors and we should not miss the opportunity.
It has recently been revealed that the EU is preparing a broad-ranging ‘Digital Services Act,’ that would allow the bloc to regulate against illegal material online, giving clearer rules for platform liability. Do you think regulation of this nature is necessary? What will the challenges be of adopting such an act?
The framework governing online content in the EU is highly fragmented. Whether addressing services like social media or activity areas like online advertising, the laws within the EU remain divergent. It is necessary for the EU to harmonise the governance of online content and address risks like online hate speech, discrimination, or disinformation. The expected Commission proposal for a Digital Services Act should be aimed at harmonising the divergent rules for online services in the Digital Single Market.
Moreover, the proposal should be aimed at incentivising large companies to tackle harmful online content beyond the voluntary self-regulating framework under which they operate now. Lastly, the Commission proposal should emphasise the need for stronger public oversight and mechanisms to address content moderation systems. A harmonised framework addressing the needs of the modern and evolving landscape of online services will also benefit innovators as barriers to access the European market will diminish.
Member states have recently submitted their risk assessment reports for the future of Europe’s 5G network. What do you think are the most pressing issues when thinking about the future of Europe’s 5G infrastructure? What are the challenges and solutions?
5G networks will be the cornerstone of the European economy and society in the future. In 2025, worldwide 5G revenues are expected to amount to €225 billion. 5G networks will seamlessly connect billions of devices, and enable the transmission of massive amounts of data between and among connected devices. The most pressing issues in deploying 5G networks at scale within the EU relate to ensuring the stability and security of those networks from potential harms.
5G networks not only enable steadier and faster connection for our devices but also connect critical infrastructure with central administration. The connection of critical services in key areas like health, finance, transportation, and resource management through 5G networks will also increase the vulnerability of those core functions to potential threats and cyber-attacks. It is imperative for the EU to concentrate its efforts in completing a coordinated European risk assessment on 5G networks.
Through the EU Cybersecurity Agency (ENISA), the EU should complete the threat landscape and the Union-wide exposures to risk related to infrastructure extending to the entire digital ecosystem. Moreover, the Commission working alongside with ENISA and member states should develop minimum common requirements to ensure higher network security.
What are your priorities going into the next mandate with the Science and Technology Options Assessment body, which you chair?
As the chair, what I can envisage is that STOA will focus a lot on green technologies and how we can use them to control environmental damage. The results of the European elections showed that citizens put environmental protection as a high priority and therefore we have to listen to their message and grasp their concerns.
We will also focus on issues such as the AI development, how to use technology for humanitarian aid and assistance, bioethics and gene editing and all the technologies that will impact our future. We will also work against fake news through our STOA creation: the European Science Media Hub (ESMH). As digital technologies advance, STOA through the ESMH is dealing with various issues from combating fake news to Virality and how it impacts politics.
[Edited by Zoran Radosavljevic]