It is a great pleasure for me to take the floor today, as part of the AI summit organised jointly by the European Parliament and the European Economic and Social Committee. We do appreciate this fruitful cooperation between our two institutions.
Artificial Intelligence is and will remain a long-term priority for our Committee. We started our work on AI back in 2017 and adopted our first opinion only a couple of months after the European Parliament's first report on Civil Law Rules on Robotics.
There is no need to stress how important Artificial Intelligence is for European citizens and businesses. Especially in the context of recovery from the COVID-19 crisis. AI is a powerful tool which can manage the complexity of the global green and digital transition we are facing. According to a recent study from the World Economic Forum, every 1 % of additional efficiency in energy demand creates $ 1.3 trillion in value between 2020 and 2050.
The COVID-19 crisis has significantly accelerated AI adoption by businesses. Before the crisis, worldwide spending on AI systems had already increased over 40 % on a year-on-year basis! This significant increase is nevertheless not the same for big companies and SMEs.
Micro companies in particular are currently at the centre of our attention. We have recently published a study [Group I study] on the uptake of AI in micro companies and we will adopt our own-initiative opinion on this topic in January next year.
One of the most pressing challenges, faced by 58 % of micro companies and SMEs, is the recruitment of new staff with the right AI skills.
Another major external obstacle to AI adoption – expressed in our study by 34 % of micro-SMEs – is the current level of uncertainty concerning liability rules for damage caused by AI. We welcome the recently launched public consultation on adapting liability rules to the digital age and AI. Unified rules across Member States are crucial to provide legal certainty and trust.
This brings me inevitably to the question of trust. Europeans do not think that the current rules efficiently regulate and oversee AI. Around 56 % of Europeans have low trust in authorities to exert effective control over AI (in 2020). The main concern is that AI can lead to unfair discrimination based on individual characteristics or social categories. This is why the recently proposed AI Act – at the centre of our discussions today – is urgently needed!
We do support the fact that the AI Act puts health, safety and fundamental rights at its centre and is global in scope. However, we also see areas for improvement of the text in its current form.
I would like to share four of the main recommendations expressed in our opinion on the AI Act adopted in September:
First, the "human-in-command" approach needs to be strongly advocated. The AI Act should make clear: AI objective should be augmenting human decision-making and human intelligence, rather than replacing it.
Second, we call for complaints and redress mechanism for citizens that have suffered harm from any AI system. Every citizen must have right to challenge decisions taken entirely by an algorithm.
Third, we welcome the governance structure – AI Board – but we recommend the AI Board to hold regular obligatory exchanges of view with wider society, including the social partners and NGOs.
Fourth, we do support a ban on the use of AI for automated biometric recognition in publicly and privately accessible spaces. As also stated in one of the morning session by the European Data Protection Supervisor's office. We are following with great attention the lively debate on this topic in the European Parliament.
The EESC is convinced that any future-proof and innovation-oriented regulatory framework must address potential concerns raised by certain uses of AI technologies. That is the right path to take.
We are convinced that the proposed AI Act and the coordinated plan on AI are key to building Europe's competitiveness at global level and confidence amongst stakeholders. We are strongly committed to working towards this common objective to make the EU the global driving force for human-centric, inclusive, safe, sustainable and trustworthy AI.
Thank you for your attention.