Artificial intelligence has dominated in China and the United States but the European Union is getting serious about this technology. Innovations in AI are accelerating and Europe is adapting with their new proposals around artificial intelligence.
Europe accounts for a share of technology companies in the AI space but with China and the US leading the pack, the EU is developing a strategic AI plan¹ for the next decade.
The EU 2018 AI strategy according to Ursula von der Leyen, EU Commission President, will define AI applications based on human rights and enhancing public trust.
Negative implications of AI including black-box systems and ethical concerns are driving the EU’s plan to develop effective proposals. Ursula von der Leyen points out that AI should facilitate human socialization in a transparent, ethical and responsible manner.
Overview of AI Regulation by the European Union
Trust from consumers is critical to the success of AI regulation in Europe and the EU commission is encouraging public participation.
Regulating AI comes with controversies and Europe wants to avoid the Chinese situation of extreme surveillance that undermines privacy. The same applies to capitalist surveillance in the US where consumer privacy comes into question.
Nevertheless, the EU will implement an AI strategy that resonates with public trust through ethics and accountability. The European Parliament and EU governments will make these decisions and develop a digital strategy² for the future.
The EU will spend $22B for AI policies across Europe in the next 10 years by promoting collaboration from citizens and businesses. According to the EU, creating reasonable AI policies in health, manufacturing and supply chain will enable the adoption of this technology and reduce risks.
The potential of AI risks³ is real and the EU is drafting policy measures that will encourage accountability in the use of AI. Autonomous vehicles fall under this category as the EU aims to protect citizens at all levels.
AI Legislation Process has started in Europe
The EU Parliament is working on AI policies with member nations meeting to chart the way forward. Europe needs a clear data strategy and with digital transformation accelerating, the legislation process will empower European companies to adopt AI and compete in the global market.
The EU anti-trust regulations⁴ require adjustments since adopting an AI strategy means transparency from tech companies about their data. Other areas including gender discrimination remain vital in the legislation given the example of the Amazon AI scandal, which favored males over females during recruitment.
Data storage is critical as the data economy grows and Europe is working on measures to ensure enough storage capabilities. This requires investments in infrastructure and the EU will spend €6B to accommodate the digital expansion.
Access to data comes second and explores public consumption of data by instilling ethics and accountability. Thirdly, the EU laws will address data sharing which comes with challenges such as privacy invasion and third party involvements.
The AI legislation process has started in the EU. Member nations are brainstorming on these proposals, and analysts expect the new laws to be ratified by the end of 2020.
EU Artificial Intelligence Proposals
Measuring the benefits and drawbacks of adopting AI by the EU is the first step towards creating open and fair policies. Unlike some countries where AI has become a spying tool, the EU wants to adopt an AI that meets the interests of all.
Here are some core values the EU is using to develop new AI regulation for member states:
1. Human and Inclusive AI
The success of AI in Europe depends on adopting a technology based on the social needs and expectations of the public. Discrimination cases about AI abound and the EU wants an AI system that will harness the potential of people and not hold them back. Human expression is essential and using this to create EU policies will enable positive applications of artificial intelligence.
The AI adopted must benefit all people regardless of their background. According to the EU, artificial intelligence should bring people together and addressing social challenges such as crime and corruption. Such an AI will benefit society and not a few individuals.
The Pope⁵ joined the debate about Europe using a fair AI system that fosters cohesion among people. According to the Vatican, inclusive technology will create a strong fair society where everybody is accountable for his or her actions.
2. Accountable and Explainable AI
The EU proposes the use of accountable and explainable AI with autonomous vehicles falling under this category. Self-driving cars have caused controversies in the past about accidents and the EU wants the AI technology used to be accountable.
Respecting human rights is critical for any AI to be accountable and the EU shares similar views. Since AI is a black box⁶, the EU through its committees is looking for modalities of control risk areas from AI.
AI systems contradict human choices and with the self-driving case, this technology should be transparent. At the same time, driverless cars contradict direction and with the trolley problem becoming popular, the EU wants the driverless car’s issues resolved.
An explainable AI technology should complement humans and not contradict them. Explainable AI will create a culture of accountability where technology serves human interests and attracts monitoring.
3. Justice and Human Dignity
What is AI technology without human dignity and justice? The EU wants AI technology, which respects the rule of law and abides with our moral and ethical values.
The driverless car problem mentioned earlier suits this example because of passengers and those developing the technology. A driverless car that accidentally rams into a grocery store causes harm to innocent people and the EU wants common ground on this issue.
For instance, those developing AI should be held accountable⁷ for their actions in pursuit of social justice. Developers of artificial intelligence according to the new EU rules should create safe solutions that observe human dignity.
Every human needs respect and AI technology that meets this criterion promotes social justice. Bias from AI algorithms poses risks to human socialization such as unfair hiring practices and racial targeting.
4. Reliability of AI Technology
A reliable AI technology according to the EU will account for the needs of all citizens such as medical fields and self-driving cars. The medical field using AI will offer benefits for Europe including cancer treatments; accurate scanning and detecting diseases early such as heart conditions.
The EU plans to implement an AI technology that will safeguard the health needs of all people. AI in surgery applies here as the technology can perform risky surgeries that humans take long to accomplish.
Driverless cars should act per human directions and incidents of a crash and stopping without consent highlight problem areas that the EU policy seeks to address. In these scenarios, the AI technology does not meet the reliability expectation and goes against expectations. An AI policy that operates in line with human interests will benefit citizens of Europe by meeting their interests.
5. Individual Privacy
The privacy issue remains a hot topic and with the emergence of AI, the EU policy proposals want a technology that respects privacy. The surveillance measures implemented in China could have concerns for privacy because of risks posed when third parties have the power of your data. Europe does not want a similar situation and the EU privacy measures⁸ around AI seek to close this gap.
Back to driverless cars, cameras installed on them record passenger activity and such information needs control to avoid damaging implications. An effective AI regulation for Europe should revolve around driverless cars and other areas where data recording occurs. Respecting personal information and privacy is critical for the AI policy proposals prepared by the EU.
Facial recognition technology has caused debates in Europe with member states pointing to the risks associated with this technology. The use of facial recognition in crowds to single out a person has drawn criticisms because of infringing on personal privacy. EU countries are divided on this issue with most opposing surveillance but Germany and others seem okay with facial recognition.
EU Commission and EU Parliament
The EU Commission and the EU Parliament are two main bodies spearheading the adoption of AI in Europe. Both bodies are working together through consultations of member states, businesses and the public on the best ways to adopt AI.
Formulation of rules governing AI falls under the mandate of the EU Commission, which coordinates policy measures, gaps and best practices. Adopting AI in Europe⁹ means looking at all areas of human socialization and the EU Commission handles this mandate.
Oversight roles from the EU Parliament ensure that AI policies adopted agree with regulations from all member states. The EU Parliament should engage with member states on all areas of AI adoption and this means ratification during the passage of the regulations.
Establishing standards matters to EU adoption of AI and the EU Commission members are working on all governing principles. For example, the EU plans to introduce AI Certification Standards for member states to enable a fair and accountable environment for the adoption of AI.
The Future of AI in Europe
Artificial intelligence for many countries in Europe offers them an opportunity to join the global race for digital transformation and value addition. Despite the risks posed by AI, Europe needs this chance to join global players in using AI for human good.
From transportation, healthcare, retail, and manufacturing, AI will transform these industries and turn Europe into a technology player on the global scene. Adopting these regulations poses hurdles that the EU should think about and come out with a good plan that serves the interests of humanity.
The EU Commission plans to increase funding for AI and boosting research and development around Europe. Testing centers for AI will create hubs that businesses, governments, and the public will participate to learn and use AI for the good of Europe.
¹EU AI Strategic Plan, ²European Digital Strategy, ³Potential of AI Risks, ⁴EU Anti-Trust Regulations, ⁵Vatican AI Policy Recommendations, ⁶AI is a Black Box, ⁷AI should be Accountable, ⁸EU privacy measures around AI, ⁹AI in Europe