Challenge 6: Accompanying the transformation

Artificial Intelligence (AI) is both a technological and social innovation, since it brings with it all the benefits and complexities that can radically transform society, including the public sector. An innovation, therefore, that can contribute to improving the quality of the services offered and to reinforcing the relationship of trust between administration and citizen.

The opportunities offered by the AI concern both the increase in efficiency of administration operations and user satisfaction. To exploit them to the fullest and to ensure that citizens fully understand their advantages and potential, it is also necessary to deal with issues concerning governance, the use of new technologies and the ability to manage data.

An aspect not to be underestimated in our country is related to the existence of a fullbodied role of “intermediaries” in the relationship between citizens/businesses and the Public Administration, combined with a culture of “delegation” that often introduces a real barrier in the relationship between users and institutions.

In this sense, it will be desirable to invest heavily in the cultural change necessary to create a substrate on which to reposition, in the key of simplification and use of digital, the new relationship between citizens/businesses and the Public Administration.

Furthermore, the transformation process we are witnessing involves the creation of a culture within Public Administration that includes capacity building activities, both with respect to the presence of a leadership that promotes the use of artificial intelligence, as well as the capacity of public officials to implement them.

Within the scope of public services, AI can be used to optimise the internal resources of PA to increase the use of online services, supporting, for example, a series of activities such as:

• completing complex tasks;
• dispatching citizens’ requests and answering their questions;
• effectively managing large amounts of data;
• combining information from different datasets;
• providing faster answers based on predictive scenarios;
• automating repetitive processes;
• analysing data that includes text/audio and video information [1].

Finally, the integration of AI can contribute to increasing the capacity of public employees, as a tool to support decision-making and without ever replacing human judgment. The immediately perceptible benefit, together with the possibility of having systems that learn to accompany decisions in an accurate and personalised way, is the possibility of saving time for employees who can dedicate themselves to more specialised activities or that require greater creativity and empathy. In this way, services become more efficient, relations with citizens are improved and the level of trust in institutions is increased.

The introduction of AI in people’s lives requires the design of processes that facilitate the understanding and acceptance of technologies by the user, not only through the use of experimentation but also through collaboration mechanisms that allow citizens to participate in the design of AI platforms.

Thanks to the co-creation approach, as happens in design thinking, users perceive technology as their own and show a greater propensity to use it. Moreover, where issues or problems in its use are found, citizens show a greater propensity to actively participate in their solution [2].

In facilitating the vicinity and engagement of citizens towards new AI-based public services, design itself plays a key role. In fact, it represents the meeting point between technology and people.

Designers will have to design interfaces that do not just mimic human actions, since this mechanism can generate alienation, but that are able to establish a relationship of trust with citizens, using a language that is understandable and that puts them at ease [3].

The challenge will be to build flexible systems able to provide answers that adapt to the user’s contingent needs, thus ensuring better and more efficient services. A peculiar characteristic of AI is indeed that of correlating continuously evolving data coming from multiple sources and extracting dynamic response models from it.

Another area on which the designers will have to focus will be the design of AI systems able to anticipate the needs of citizens without having an invasive approach that could compromise the user experience.

Another crucial element for introducing AI in a structured manner in the administration concerns the ability to manage data and to exploit the great wealth of information that PA possesses, facilitating not only interoperability, but also transparency and reliability.

In light of this, it is desirable that the application of AI technologies to public administration aims at adopting shared ontologies in line with the internal organisation of PA and with the types of services to be provided, developing controlled vocabularies able to interpret and interoperate the databases of national interest to the fullest [4].

In this regard, knowledge, representation, and self-learning systems can be a valuable aid in increasing the accountability of the models. The adoption of collaborative methods can further ensure that the models adopted are compatible with PA and remain consistent with the regulatory framework.

In addition to the potential described above, some criticalities in the adoption of AI in the public sector can be identified: in general, AI systems can be implemented successfully only if high data quality is guaranteed.

In terms of governance, the transformation process we are witnessing also involves the evolution of relations between public and private players.

Benefiting from AI in public services does not necessarily mean developing new solutions from scratch. On the contrary, it is possible to look at what has already been adopted by other governments, or draw on technologies already established on the market.

Area of collaboration between the public and private sectors is that of procurement. In this sense, AgID, for example, has recently started initiated comparison and experimentation of new scenarios for the dissemination of PCP (pre commercial procurement).

The program deals with issues of significant social impact and public innovation: from autism to protection from environmental risks, to food safety and quality, as well as innovative technological solutions applied to healthcare and e-government. Not only large companies but also start-ups, small businesses and venture capitalists have the opportunity to present innovative ideas and proposals. The PCP is therefore a fertile ground for experimentation and research aimed at meeting social needs even with innovative tools related to AI. An example in this sense is the “Technologies for Autism” contract aimed at identifying Virtual Reality and Augmented Reality technologies typified for people with an autism spectrum condition (ASC).

Footnotes

[1]	Hila Mehr, “Artificial Intelligence for Citizen Services and Government”, HARVARD Kennedy School ‐ ASH Center for Democratic Governance and Innovation, 2017

[2]	Medium, “The role of design in collaborative AI”, 4-11-2017

[3]	Medium, “Human Design for Artificial Intelligence and Agents”, 19-10-2017

[4]	Ref. http://pianotriennale-ict.readthedocs.io/it/latest/doc/04_infrastrutture-immateriali.html.

[5]	Italy is first in the ranking of EU countries for capacity to implement pre-commercial procurement