Summary:

• AI should be the EU’s true moonshot. We should not only develop safe AI but using AI to ensure the safety and security of Europeans. The overall narrative should be changed by showing examples of using AI for citizens’ benefit.
• If Europe cannot fully compete in the development of AI in all areas, it should build on quality, rather than quantity of R&D and support a competitive AI ecosystem. The regional specialization should be embraced and the EU-wide AI hubs focused on such specific areas should be supported. DARPA style research calls should be set up to attract talents from outside of the EU.
• The EU should draw strict lines to prevent misuse of AI, similar to how the GDPR set clear rules for the use of technology without compromises. However, the legislation should be enacted only to answer market demand.
• New AI regulation should set clear rules to create legal certainty and support the development of AI, not to hinder innovations and undermine Europe’s position in the AI race.
• Europe should focus on leadership in AI applications in line with fundamental rights and values. Its main competitive advantage should be a strong emphasis on unbiased, societally benefit and pro-democracy oriented AI applications, especially towards developing countries.
• Europe should embrace AI deployment to companies (esp. SMEs) ahead of China and the US, support the potential future champions and direct financial support towards this goal. It should be the ultimate tool for convergence of EU economies.
• All Europeans should have an equal possibility to gain benefits from AI and automation. It should be one of the goals for the planned European Future Fund.

Policy Recommendations – Institute for Politics and Society, January 2020

Full text in PDF under the link bellow.