Tuesday

21st Nov 2017

EU unsure how to 'make most' of AI

The 28 leaders of the EU's member states agreed last week that by "early 2018" there should be a "European approach to artificial intelligence".

They also said in the official summit conclusions that the EU needed "a sense of urgency to address emerging trends" such as artificial intelligence (AI).

Thank you for reading EUobserver!

Subscribe now for a 30 day free trial.

  1. €150 per year
  2. or €15 per month
  3. Cancel anytime

EUobserver is an independent, not-for-profit news organization that publishes daily news reports, analysis, and investigations from Brussels and the EU member states. We are an indispensable news source for anyone who wants to know what is going on in the EU.

We are mainly funded by advertising and subscription revenues. As advertising revenues are falling fast, we depend on subscription revenues to support our journalism.

For group, corporate or student subscriptions, please contact us. See also our full Terms of Use.

If you already have an account click here to login.

  • Noel Sharkey: 'Over the years I've seen AI systems that then just become normalised and they're not AI anymore' (Photo: Kosciuszko Institute)

This reference in the summit conclusions was sign that politicians attach great importance to AI as a means to maintain a competitive edge.

Perhaps one of the most clear and somewhat ominous expressions of this was a recent quote attributed to Russian president Vladimir Putin.

"Artificial intelligence is the future, not only for Russia, but for all humankind," Putin said last month according to RT, a Russia-funded TV network.

"It comes with colossal opportunities, but also threats that are difficult to predict. Whoever becomes the leader in this sphere will become the ruler of the world," he added.

Later in September, France's new president also explicitly referred to AI, in a speech in which he laid out his vision for Europe.

"Why can we not start a 'disruptive innovation agency' and launch a joint artificial intelligence programme, which would make Europe a driver of global growth?" Emmanuel Macron asked.

He said Europe should create this European agency in the next two years "in the same vein" as the US Defense Advanced Research Projects Agency (Darpa), which helped kickstart the birth of the internet.

"Such an agency would make Europe an innovator and not a follower," said Macron.

After last week's EU summit, Finnish prime minister Juha Sipila said in a press release that Europe "cannot afford to lose the best ideas to competitors and other markets".

"We must start the work now," said Sipila, whose country claimed credit for the reference to AI in the summit conclusions, and is the home to several AI startups.

What is artificial intelligence?

There is no commonly-agreed definition of AI, which sometimes is mixed up with other concepts, like machine learning, self-learning or autonomous systems.

In fact, the concept has evolved over time, said Noel Sharkey, a computer scientist and emeritus professor of artificial intelligence and robotics at the University of Sheffield.

"AI is actually the name of a field," Sharkey told EUobserver recently at the Cybersec conference in Krakow, Poland.

"Over the years I've seen AI systems that then just become normalised and they're not AI anymore, they're part of software engineering," he added.

Voice recognition was once considered AI, but has quickly become a mainstream feature of many applications.

The phrase 'artificial intelligence' has been around since 1956. The first mention of it by the European Commission is thirty years old.

"Artificial intelligence or ... adaptive intelligence could lead to the manufacturing of industrial robots which could literally see what they are doing whereas today they blindly accomplish pre-programmed tasks," the EU's executive said in a 1987 press release.

In 2017, AI is seen as a crucial technology to achieve self-driving cars, but it could also help doctors to find patterns in symptoms based on large data sets.

The phenomenon is linked to that of 'big data', which in turn was made possible in recent years due to increased data processing capability in the 'cloud'.

Reducing traffic deaths and energy waste

The European Commission's digital single market strategy paper from 2015 did not mention artificial intelligence specifically, but its mid-term review two years later said the EU should be "in a leading position in the development of artificial intelligence technologies, platforms, and applications".

"Use of artificial intelligence in different technological solutions can lead, for example, to fewer fatalities on roads, smarter use of resources such as energy and water, less pesticide use on farms, and a more competitive manufacturing sector," the commission said in its report from May 2017.

The EU's executive body has mostly focused on supporting AI research through grants, but also on legislation needed to enable AI development.

The commission is said to be spending around €1 billion in 2014-2020 on research projects on robotics and AI, more than two-thirds of that through a public-private partnership.

In the legislative arena, the commission's proposal on the free flow of non-personal data, presented last month, aims to "unlock" the "potential" of AI and other digital technologies.

In their summit conclusions, EU leaders called on the EU institutions to agree on a compromise on the free flow of data proposal by June 2018 - which would be relatively quick for an EU regulation.

But there have been calls on the EU to look at other aspects of AI.

Ethics

Two EU institutions, representing European citizens and civil society, have produced thought-provoking reports on the subject this year.

The European Parliament adopted a text last January which said that "ultimately there is a possibility that in the long-term, AI could surpass human intellectual capacity".

The non-binding report, supported by 396 MEPs and rejected by 123, asked the commission to "consider the designation of a European Agency for Robotics and Artificial Intelligence".

It said the EU "could play an essential role in establishing basic ethical principles to be respected in the development, programming and use of robots and AI".

However, it also claimed that "in the scenario where a robot can take autonomous decisions, the traditional rules will not suffice to give rise to legal liability for damage caused by a robot".

The MEPs called on the commission to "explore, analyse and consider the implications" of "creating a specific legal status for robots in the long run, so that at least the most sophisticated autonomous robots could be established as having the status of electronic persons responsible for making good any damage they may cause".

Legal 'personhood' is a bad idea

While emeritus professor Sharkey liked "the core" of the parliament's document, he rejected the idea of giving legal personhood to autonomous robots or AI systems, like has been done with corporations.

"The idea of doing that with a robot means you are taking away responsibility and accountability from humans," he said.

The AI expert said that the question of who is responsible when a self-learning system makes an autonomous decision was "the wrong way to frame it".

"You hold people liable and if they have developed a machine that does these things and they can't understand it, that's their problem, not the machine's," he said.

Sharkey imagined what would happen if a robot or machine would be liable for its actions.

"The machine would have to accumulate money itself so they could pay out liability. But what if it's criminal?"

Humans in command

His views were shared by the European Economic and Social Committee (EESC), which represents civil society groups.

"The EESC opposes the introduction of a form of legal personality for robots or AI. This would hollow out the preventive remedial effect of liability law; a risk of moral hazard arises in both the development and use of AI and it creates opportunities for abuse," the EESC said in a report adopted in May 2017.

It said that a comparison with companies was "misplaced", because those are ultimately always headed by a natural person.

The deliberative body said the EU should take "a human-in-command approach to AI", requiring that "machines remain machines and people retain control over these machines at all times".

The text was supported by 159 EESC members – only three voted against it.

In a response to the parliament's report, the commission said it agreed it was "important to examine whether and how to adapt civil law liability rules to the needs of the digital economy".

It has asked for input through two public consultations and is still assessing the results.

The "European approach" asked for by EU leaders is expected to be non-legislative, like a strategy paper.

On Tuesday (24 October), the commission presented its annual work programme for 2018. It said the commission will "look to make the most of artificial intelligence", but in the annex of initiatives did not announce any specific measures on AI or liability.

However, if an issue is not in the work programme, that does not automatically mean that it is excluded from being considered.

Sex toys and smart robots: Who's liable?

Sex toys are being connected to the internet ad they can be hacked. Is that rape? And who would be responsible for misdeeds of autonomous robots?

Supercomputing lag could prompt EU brain drain

“We are not in the top-10 or the top-five in the world when it comes to high-performance computing but we have the potential to do it," says EU digital commissioner Mariya Gabriel.

Stakeholders' Highlights

  1. Bio-Based IndustriesBio-Based Industries: European Growth is in Our Nature!
  2. Dialogue PlatformErdogan's Most Vulnerable Victims: Women and Children
  3. UNICEFEuropean Parliament Marks World Children's Day by Launching Dialogue With Children
  4. European Jewish CongressAntisemitism in Europe Today: Is It Still a Threat to Free and Open Society?
  5. Counter BalanceNew Report: Juncker Plan Backs Billions in Fossil Fuels and Carbon-Heavy Infrastructure
  6. Nordic Council of MinistersNordic countries prioritise fossil fuel subsidy reform
  7. Mission of China to the EUNew era for China brings new opportunities to all
  8. ACCASmall and Medium Sized Practices Must 'Offer the Whole Package'
  9. UNICEFAhead of the African Union - EU Summit, Survey Highlights Impact of Conflict on Education
  10. Nordic Council of MinistersNordic Council Calls for Closer Co-Operation on Foreign Policy
  11. Swedish EnterprisesTrilogue Negotiations - Striking the Balance Between Transparency and Efficiency
  12. Access EuropeProspects for US-EU Relations Under the Trump Administration - 28 November 2017

Stakeholders' Highlights

  1. Nordic Council of MinistersSustainable Growth the Nordic Way: Climate Solutions for a Sustainable Future
  2. EU2017EEHow Data Fuels Estonia's Economy
  3. Mission of China to the EUChina and EU Step Up Water Management Cooperation
  4. CECEMachinery Industry Calls for Joint EU Approach to Develop Digital Construction Sector
  5. Nordic Council of MinistersMale Business Leaders Gather in Copenhagen to Advance Gender Equality
  6. EnelNo ETS Deal Means It Can Still Be Strengthened
  7. EU2017EEEstonia Anticipates More Digital Cooperation With Sweden
  8. Mission of China to the EUChina Launches Campaign to Protect IPR of Foreign Companies
  9. European Jewish CongressEJC Condemns Attacks on Ruta Vanagaite and the Shredding of Her Books in Lithuania
  10. Bio-Based IndustriesDiscover the Future of the Bio-Based Economy. Register Now for the BBI Stakeholder Forum!
  11. European Free AllianceWelcome Catalonia!
  12. UNICEFGrowing Number of Unaccompanied Refugee Children in Greece in Need of Shelter