A new path for AI (education) within the public sector

On 15 September 2021 I will be participating in the panel on AI Education and Skills at the High-Level Conference on AI: From Ambition to Action. This is a great opportunity to discuss education for/with Artificial Intelligence (AI) and, in particular, its implications and impact in the public sector from the perspective of European values and competitive advantages.

AI has indeed the potential to contribute in addressing some of the biggest challenges of our contemporary society. At the same time, it can also empower transparent governance, the reduction of the many existing digital divides, and hold up the process of civil society empowerment. Indeed, following the criteria highlighted in the proposal for a Regulation on AI laid down by the European Commission, specific applications of AI systems operating in private and public services, if not used with care, may create conditions of high risk. This is why it is necessary that the users of AI systems, are adequately informed (and trained I would add) while appropriate human oversight can minimise such risks. It is essential to have a skilled and informed workforce, able not only to understand new technologies but, more importantly, to be able to design and use them in a way that can provide enhanced services and solutions. Therefore, educational initiatives for public institutions related to the uptake and use of AI can create timely and strategic opportunities.

The Recovery Plan structured by the European Commission had among its priorities to modernize Europe's public and private infrastructures. In particular, the EU Digital Agenda has the enforcement of the digitalization of public services as one of its main pillars, improving the competencies of European public servants is included. The data highlighted by DESI - Digital Economy and Society Index 2020, shows clearly how it is necessary to expand the knowledge of Europeans and their administrations on the technological and digital level.

But mere knowledge is not enough, nor is it a matter of filling positions in public bodies with programming experts. What is needed is the education of “functional specialists”. An educated functional specialist in AI and public services would be able to conduct (i) the management of AI-based public services, (ii) the use of AI in public administration’s operations, (iii) the governance of AI in the public sector and for the delivery of public services to society, and (iv) the design - and eventual implementation- of AI-based services within public institutions built on European values and able to display the advantages of the European approach to public services.

As far as the technological level of public services is concerned, the departing point is relatively low although the context is favorable. Using an upward trend, accelerated by the pandemic, most European administrations have improved their online services, both in terms of quality and use. This qualitative leap is evidenced by users' activities: in 2019, it was recorded that 67% of digital users complete and send administrative documents via forms and digital platforms, and 64% of users had interactions online with the public services. Despite this, smart e-gov services are infrequently used by citizens in many European countries. The reason relates to the fact that even if the demand for digital services is constantly increasing, the offer is often inadequate, or even absent. A sense of distrust and frustration may naturally develop on the part of users who continue to see the public institutions as a structure of inefficient and obsolete procedures. However, following the pandemic, expectations about the digital future of public administrations have risen considerably. Their modernization and the strengthening of digital governance could provide citizens with a more agile, transparent, open, and understandable system.

Precisely for this reason, with colleagues from four leading European universities -Universidad Politécnica de Madrid (UPM), Politecnico of Milano (PoliMi), Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), Tallinn University of Technology (TalTech)-  of Tallinn we have launched what we expect will become a world-class master’s program on the use of AI in public service delivery, the Master in Artificial Intelligence for Public Services (AI4Gov). Our shared vision -that received the support of the European Commission Connecting Europe Facility- aims at boosting the public sector’s capacity in managing, understanding, and developing AI-related projects and services in the certainty that they will improve the provision, effectiveness, and value of public services throughout European society. Actual figures from registration of prospective students far exceed available places, proving the demand for increasing AI skills of personnel involved in crafting more efficient public services.

But we should not just focus on education but rather on its key role in improving the delivery of public services. The real goal is the development of an ecosystem AI-talent-cantered in public services, conducive for creating new opportunities within public institutions. For this, and following the proven approach of innovation ecosystems, relevant stakeholders are called to participate and co-create these opportunities, be public bodies, universities, students, innovators, intra/entrepreneurs, startups, non-governmental organizations, and private sector businesses in a context of public policies, regulations, infrastructures, and culture.

The position of Europe here is unique. Enabled by our democratic values and human rights, we can confidently explore public service possibilities that empower and strengthen civil society in the knowledge that this will only make our countries stronger and more resilient. At the same time,  we are well aware that human dignity and privacy are values that are worthy of protection from the start. These are the challenges and the true contribution of future “AI in public services” functional specialists.



Iesniedzis azamat ABDOULLAEV Pr, 13/09/2021 - 11:37

The Old-Style AI master/PhD educational programs and courses miseducate young people and shouldn't be funded by the EU. Most popular ML/AI courses are most obsolete and useless for future prospects:  

  • AI For Everyone: DeepLearning.AI.
  • Machine Learning: Stanford University.
  • IBM Applied AI: IBM.
  • AI Foundations for Everyone: IBM.
  • Mathematics for Machine Learning: Imperial College London.
  • Introduction to Artificial Intelligence (AI): IBM.