Ladies and gentlemen,
It is a pleasure for me to be here at the AI France Summit 2020 and open this roundtable on today's and tomorrow's AI leaders.
I would like to thank Cédric O, Secretary of State for the Digital Economy, for inviting me.
France clearly plays a leading role when it comes to AI. The talent gathered in this room proves it.
Vibrant startups, leading industries, brilliant researchers, innovative businesses. And a strong political drive. France was one of the first countries to present a comprehensive AI strategy back in March 2018, with a clear European dimension. It was important for our work within the Commission - I was Commissioner for Justice at that time and remember this well - and it is still shaping our European vision for digital today.
Let me tell you more about it.
[The European digital way: people and their rights first]
The new Commission has just presented its digital blueprint for the five coming years, along with specific strategies on AI and data. My role, as the Vice-President for values, was to ensure that fundamental rights are embedded in these plans. Because, at the end of the day, this is not only about technological progress, but first and foremost about human progress.
Let me tell you from the start - I strongly disagree with those who argue that innovation and fundamental rights cannot go hand in hand. Europe has shown one is actually supporting the other.
Diversity, equality, right to private life, consumer protection - these are not a luxury, they are a necessity in the digital world.
Because these values, these rights are part of our European DNA. They are at the heart of a European model, which put technologies at the service of the people, not the other way round.
People first. Not the State first -, using AI to monitor and score people. And not laisser-faire first, where fundamental rights are marginalised.
Our European model must be based on trust. For the digital revolution to succeed, people must trust the technology, and if they trust it, they will use it. It also means better products and services. This is more responsible, more sustainable. And the only way, in my view, to mitigate a “techlash”.
The big test for this approach will be AI, which already has an increasingly important role in our daily life.
I do not have to tell you about the great potential of AI to improve our lives - or even to save lives, when you think about healthcare and how AI can help doctors predict cancer better and earlier.
You are very much aware of the benefits. And in the European Commission, with Commissioner Breton whom you know very well here, we want to make the most of these opportunities.
This is why, in the White Paper we have just presented, we put a great focus on promoting excellence and trust.
Promoting excellence, this means investing in the key ingredients of AI - the first, for me, are actually human brains. Interesting when you talk about AI, right ? Then data - the raw material for AI.
[Diversity for AI]
I start with talent. We can only reach our full potential if we use all of our talent and diversity. This morning, I presented the Commission's gender equality strategy.
Women represent more than half of the population, but only 17% of people in ICT studies and careers in the EU and only 36% of STEM graduates.
At the same time, there is a shortage of ICT experts and we know how important diverse teams are to develop the best products and services.
A real concern about AI is that will replicate or magnify bias existing in the real life. The fact, for example, that a recruitment system based on AI would exclude or demote women or minorities from its results. If you have diverse teams and strong ethical standards, you can develop more inclusive AI.
Tomorrow, I will have the opportunity to visit Willa, an incubator for women startup founders. They shared with me some interesting data: founding teams that include women outperform their all-male peers by 63%.
Diversity, this is also about working together, public sector, companies, academia, civil society. This is why we will present a European public-private partnership on AI by the end of the year.
We need to connect the dots - and research excellence centres from various regions of Europe. This is similar to what you are doing at French level, with a network of interdisciplinary institutes, here in Paris, but also in Grenoble, Nice and Toulouse. We are investing in such networks at European level.
[Investing in our future - unlocking data]
I mentioned investments. Which amounts are we talking about ?
In a context of fierce international competition, we should aim to reach at least 20 billion euros of public and private investments in AI per year by 2025.
To reach this target, we need to create the right environment to stimulate investments - and this is a strong digital single market. With common rules - instead of a patchwork of national frameworks - companies can scale up across borders. A lot has been achieved over the past years - when it comes to cybersecurity or data for example. I led the work on GDPR, we adopted rules on the free flow on non-personal data in Europe, and I negotiated data flow arrangements with important trading partners such as the US or Japan. We need to make sure that those rules are properly enforced and we need continue removing remaining barriers.
We also need to use public funding to leverage private investments. We thought of this when we proposed our next EU budget for the period 2021-2027. For the first time, we proposed a programme dedicated to digital - Digital Europe - and the most ambitious European research and innovation programme ever, Horizon Europe, with 100 billion funding. As European leaders are now in the last miles to agree on our future budget, it is essential to keep in mind what tomorrow's growth and jobs will be about.
We also need data.
AI is as good as the data feeding it.
We have a goldmine of industrial, research and public data in Europe.
We will come with measures creating incentives for data sharing between businesses, between businesses and governments, and within administrations. This means establishing practical, fair and clear rules on data access and use, which comply with European values and rights, including with the GDPR, our flagship of a modern and horizontal privacy regime which has also become a reference point for other partners in the world.
[Trust and openness]
Promoting talent, investments, data sharing - this is about our ecosystem of excellence. Let me get back to the other leg other our AI White Paper. Trust.
For people and companies to make the most of opportunities, they need to be able to trust AI systems. And today we already see that technology is not good or bad - it's what we make of it.
I am personally very concerned when I read about facial recognition towers in Hong Kong that can identify protesters, for instance.
This is why we come with very strong recommendations for an ethical approach to AI. We should be able to trust that we will get accurate results from AI, that we will be safe, our rights are respected. And because for most of the people, including me, this technology is rather complicated, we have to also ensure that the design process is based on strong ethical standards.
I think a lot can be also be achieved with more transparency.
We already have a strong legal framework in place to ensure consumer protection, to address unfair commercial practices and to protect privacy. There are also specific rules for certain sectors such as transport.
But it is likely that additional rules are needed for high-risk applications when it comes to health, justice or police for example, and that some of the existing ones, for example on liability and safety of products, need an update.
I would like to get your views on this - so please contribute to the public consultation that we have just launched.
If we work together on this and identify the right level of regulation, we can become a standard setter on AI - as we have done on privacy rules.
But this does not mean we are working in isolation. I will be in the US next week to discuss tech regulation. On safe data flows, we have reached a deal with Japan, we are close on finishing the talks with South Korea and Canada has been extremely active on the discussions on AI and ethics at international level.
You have heard politicians talking about technological sovereignty. I do not like too much the word ‘sovereignty', to be fully honest with you, because it makes me think more about the past than the future. But I fully share the drive of being assertive and not being naïve. For me it means having the freedom of choice, having European alternatives when it comes to strategic technologies. It means having the self-confidence to remain open. We need openness, cooperation and trust.
For me, this is how France, and Europe, can be a leader in AI, today and tomorrow.