Right Honourable Members of the European Parliament, Ladies and Gentlemen!
I would like to thank the European Parliament for the invitation to speak at this seminar on a topic which is at the heart of the European Commission's activities, especially those of the Directorate-General I represent, DG Connect.
Let me first state that we greatly appreciate and welcome the publication of the report of the European Parliament Legal Affairs Committee.
Given the great public attention currently devoted to robotics and artificial intelligence (AI) this report is very timely and points at some crucial issues that need to be addressed.
It highlights the opportunities robotics and AI offer and points towards the clear need for a coherent European approach and for Europe to have a strong presence and investment in this technology in order to maintain leadership in it.
Many aspects of the report are directly in line with the Digitising European Industry strategy we launched in April 2016. Our strategy identifies Robotics and Artificial intelligence as cornerstone technologies, it addresses the new legal challenges related to robots, autonomous and AI-based systems, and it highlights issues related to skills.
In the Commission, we have long recognised the importance and potential of robotics and artificial intelligence and the need for significant investment in this area.
For this reason, we have set up SPARC, the Public-Private Partnership for robotics in Europe, to develop a robotics strategy for Europe. With €700 million EU funding and, adding private investment, an overall investment of €2.8 billion, SPARC is by far the biggest civilian research programme in this area in the world.
We need this investment to stay in the lead. Because one thing is clear: Europe will only be able to shape this and future debates and introduce European values such as human dignity if Europe maintains a strong leadership role.
And don't be mistaken: international competition is fierce. Google, Amazon, Facebook – everyone is jumping onto artificial intelligence at the moment.
At the same time, Europe is in a strong position, both scientifically and commercially. For example, about a quarter of all industrial robots and half of all professional service robots are produced in Europe, and many recent breakthroughs in AI come out of European labs such as DeepMind's AI software that beat a human Go player.
Against this background, we appreciate the support and understanding from the European Parliament to maintain and reinforce these efforts and welcome the supportive message in the report.
Let me first clarify what we mean by AI and robotics:
Firstly, we have industrial robots installed on factory floors, carrying out repetitive tasks such as pick and place or transporting goods autonomously. They are programmed to achieve very specific tasks in very constrained environments and usually work behind fences with no human contact.
Increasingly, so-called collaborative robots are deployed on the shop floor which can work in close proximity of humans and do not need a security cage any longer.
A second category consists of professional service robots used outside traditional manufacturing. Typical examples include surgical robots in hospitals or milking robots on farms.
Consumer robots form the third category: they can be used for private purposes, typically at home, like vacuum cleaners, lawn mowers etc.
Finally, there are the purely software-based AI agents. Such systems are used, for example, to help doctors improve their diagnosis or in recommendation systems on shopping websites.
AI-based software, in conjunction with sophisticated sensors and connectivity, is also increasingly used to make all kinds of devices and objects around us intelligent. The most notable example in this context is probably the self-driving car.
While many of these robots and AI systems are impressive and have progressed a lot recently, they are still very far from exhibiting intelligent, human-like behaviour or are indistinguishable from a human. In other words: they don't pass the Turing test yet. This futuristic vision would need a debate at a different level, including asking very profound ethical questions.
Given the impact on our society and economy, we are very serious about the issues raised in the report and we have in fact been working on these issues for some time already.
But we have to be cautious and address concrete problems we are facing today and carefully assess if the current legislation is fit for purpose. For instance, now attempting to regulate human-like artificial intelligence as portrayed in Hollywood movies like "Ex Machina" or "I, robot" is probably too far-fetched and speculative.
Recently, we published our Communication on Building a European Data Economy, which includes a comprehensive section on liability and puts forward possible mitigation measures.
We now wish to engage in broad consultations with stakeholders to develop a deeper understanding of the issues and possible ways to address them.
Testing and experimentation with this technology will be important as well to gather data and gain experience. This, in turn, will then help us designing a suitable legal framework.
For this reason, our Communication also includes plans to establish cross-border corridors to test connected and automated driving.
As the report rightly points out, transparency will be crucial. For this reason, we will make a concerted effort to start a dialogue with civil society about these issues and involve it in the stakeholder consultations very closely.
Any measure proposed or considered in this context should not stifle innovation – innovation which can bring concrete benefits to citizens’ everyday lives. Rightly done, regulation can also be an enabler for innovation. The report appropriately highlights the need to ensure this.
The benefits of innovation will only be reaped when robots work in practice, and we think the report correctly identifies some issues that are critical for the deployment of this technology and should be addressed urgently. Measures related to safety, for instance, are a necessary condition for the deployment of the technology.
In the Commission, we have already investigated safety issues from various angles: we have been funding research and development with the goal of making robotics systems inherently safe, and our programme also includes a specific focus on safety certification and related standardisation.
As far as the regulatory aspect of safety is concerned, the Commission is currently evaluating some existing legislation, such as the Defective Products Liability Directive and the Machinery Directive, with a view of determining their fitness for purpose.
It is worth pointing out in this context that automated systems often operate more safely than humans. The low number of airplane accidents is also the result of planes being flown by computers most of the time. Knowing that 90% of all car accidents are caused by human error, connected and automated driving promises to bring these safety benefits also to our roads.
Coming back to the report, the various issues discussed in it merit careful and detailed analysis, taking into account the actual capabilities of today's robots and AI-based systems and the expected future developments in this respect.
Indeed, AI has overpromised in the past, and therefore any decision should be based on factual information rather than unrealistic expectations from the technology.
The digital transformation of work is another crucial element in this discussion which the report rightly identifies.
This is a topic that currently receives a lot of press attention and is a cause of great concern for citizens.
The Commission is fully aware of the challenge ahead and has already launched concrete measures to address it. In December 2016, for example, we started the Digital Jobs and Skills Coalition, which aims at equipping the workforce at large with the necessary skills to thrive in a digital workplace.
At the same time, we should not forget that robots are used in many areas with labour shortages such as healthcare, farming and even manufacturing.
Furthermore, many robots do jobs that are dull, dirty and dangerous for humans such as inspecting oil tanks or welding metal parts.
Far from replacing humans, we want robots to help humans so that they can focus on the essentials. A nurse should not have to carry bed linen around at a hospital all day. Instead, this task could be done by a robot so that the nurse can focus on the well-being of patients.
We fully agree with the report that the evolution of the labour market and the impact digitalisation has on it needs to be carefully and closely monitor. Current evidence base needs to be improved and we need to work on that.
Finally, the report highlights international cooperation. This will be important because G7 and G20 have put digital transformation and AI on their agendas. We welcome the support of the European Parliament on this particular point.
Right Honourable Members of the European Parliament, Ladies and Gentlemen! To conclude, we thank the European Parliament for this timely and comprehensive report and the support for our activities.
The issues raised and the measures proposed will need broader stakeholder consultations and an in-depth analysis of their impact and consequences. Only after these findings are available and facts have been established can we conclude how to move forward, especially as regards legislative measures.
Either towards the end of this year or at the beginning of next year, we will consolidate our thinking on this topic and summarise the results of our stakeholder consultations in a dedicated Communication. This will then clarify our position.