AI Watch: The importance of LT in national AI Strategies in Europe

In 2018, the European Commission adopted the Coordinated Plan on Artificial Intelligence that was developed together with the Member States to maximise the impact of investments at European Union (EU) and national levels, on the one hand, and to encourage synergies and cooperation across the EU, on the other hand. 


One of the key actions towards these aims was an encouragement for the Member States to develop their own national AI strategies. Most recently, the AI Watch – National strategies on Artificial Intelligence study was published, presenting an updated review of national AI strategies from the EU Member States, Norway and Switzerland. The review includes the latest overview and insights from the national AI plans in relation to supporting human capital, networking, bringing AI technologies to the market, infrastructure and regulatory aspects. AI initiatives for addressing societal challenges are also specifically addressed.

From a Language Technology (LT) perspective, the report yields various interesting results:

1.    Above all, AI Watch Open confirmed that access to public sector data as supported by ELRC continues to be a priority, as national data strategies increasingly focus on AI to foster a robust digital ecosystem for AI and to advance AI R&D. 

2.    Moreover, most of the countries are also taking several actions to stimulate the use and uptake of AI in public services. Many national AI strategies include a requirement for experimental projects to learn by doing and sharing good practices. Some national strategies mention funding programmes to support AI projects in public administration. Initiatives such as GovTech Polska and GovTech Lab Lithuania are good examples of how to boost the innovation ecosystem in the public administration and AI.

3.    Most importantly, most countries highlight language technologies in their AI strategies as key to enabling interactive dialogue systems and personal virtual assistants for personalised public services. For example, Denmark, Norway, Portugal, Slovakia, and Spain report supportive policies for natural language processing. Slovakia is developing a tool for natural language processing to accelerate the development of AI in the private sector and improve the quality of public services. Spain launched a project on the Spanish Language and AI (LEIA) to promote and enable the use of the Spanish language in the digital world. These policy efforts continue the action lines outlined in the National Plan for Advancement in Natural Language Processing. Finally, Denmark is focusing on language technologies to support ‘AI in Danish’ and, in June 2020, launched a platform displaying metadata of existing linguistic resources to facilitate the development of language technologies in Danish.

The question about the real scope and weight of Language Technologies in Europe remains. Which AI related LT projects and initiatives do currently exist in the EU Member States, Norway and Switzerland? What is the available AI funding for LT? What are major LT players? And what kind of data collection efforts/repositories for LT/AI do exist? The next meeting of the Language Resource Board (LRB) of the European Language Resource Coordination (ELRC) will hence provide a more in-depth analysis of LT in the current AI strategies, drawing on potential areas for collaboration as well as good practices, in order to support national policy makers and help the EU reinforce its position for developing AI-based LT and collecting language data under the DIGITAL programme.