AI will also bring about major changes in the public sector. We take a look at some ways government organizations can use AI to improve the way they serve and interact with citizens.

Many organizations today are turning to artificial intelligence to solve their problems, both in the private and public sectors. Industry experts sometimes dare to say that those who do nothing with AI are hopelessly left behind. Belgian IT company Smals wants to stay away from such bold statements, but that certainly doesn’t mean AI is off the radar.

AI can also play a significant role in the government context. Data analysis and automation of administration are also relevant topics in the public sector. In its latest activity report, Smals takes a closer look at some of the most promising uses of AI in the public sector.

increase productivity

Almost everyone seems to agree that AI can be a great tool to free employees from time-consuming manual work. One area where Smals sees potential for AI tools is in software development. Generative AI tools like the now well-known GitHub CoPilot convert natural language into code.

In addition to (re)writing code, it can also be used to carry out automatic software tests. In 2023, Smals will look for viable alternatives that development teams can safely use: Deploying public cloud solutions remains a delicate matter.

AI can also take on the role of an intermediary in the contact between citizens and the government. In order to be able to process a larger number of questions better, Smals is researching the use of speech recognition intent detectionTechnology calls can be routed more efficiently. AI can take over simple questions, but AI can also help with suggested answers for more complex questions. New developments in generative AI will bring new opportunities.

Advanced analytics

The added value of AI can also be felt in data analysis. The public sector has a wealth of data. The knowledge, techniques and methods for its analysis are refined and expanded every year. For example, geographic analysis in government databases can provide useful information in a variety of contexts, from fraud prevention to impact assessments and medical guidelines.

Smals believes that by subsequently incorporating this geographic information into more advanced analytical techniques such as network analysis, the level of data analysis can be increased. Those network analysis techniques that can be used to uncover fraudulent connections between entities, for example, can in turn be extended by machine learning techniques.

While machine learning is mainly used to look for patterns in historical data to predict the likelihood of something happening in the future, causal AI goes one step further and looks for causal relationships in the data. Decisions or suggestions based on data analyzes should thus become easier to explain and check. Causal AI is still a relatively young field of research that can lead to major changes in data analysis.

Say it in natural language

Natural Language Processing, the processing of natural language by algorithms, is perhaps one of the areas where AI has made the most spectacular advances. As a result, chatbots have more and more application possibilities and also the quality of the output is visibly better than before, thanks to the availability of large language models such as GPT-3.5 and GPT-4, which have been causing a stir for several months.

Smals is experimenting with this in social security, among other things, to enable people to make statements by voice. The government can also allow people to ask questions question-answer systems B. via websites of authorities or internal data sources. Care must be taken to ensure that the answers are correct.

However, there are still some areas that need work. For example, the language models used must be further refined in order to better understand the technical jargon. develop one distributed cloudwhere AI solutions from major cloud providers run on their own infrastructure, allows you to create your own translation apps with a better guarantee that personal data is processed in accordance with GDPR legislation.

Chatbots have more and more uses and the quality of the output is visibly better than before.

Privacy first

Technology is there to make citizens’ lives easier, but technology should not come at the expense of their privacy. Cryptography is the science that deals with privacy. A new trend in cloud computing called Confidential Calculation applies the principles of cryptography to the cloud. Personal data could be stored in a public cloud without the provider having access to the data.

But what if the key to privacy lies within the data itself? Where possible, Smals uses already advanced pseudonymization to hide, for example, national register numbers or medical data of citizens under a veil of anonymity. This makes the data more useful for building machine learning models.

This principle can go even further. Synthetic data is “fictitious” data that replaces confidential data in a data set, with as little impact on the statistical characteristics as possible. The extent to which synthetic data can already be used to analyze public databases and how AI can be used to generate this data promises to become an interesting research field in 2023.

API economics

Artificial intelligence can only be worthwhile if the right technological architecture is in place. Smals believes that AI can be a driver for the development of an “API economy” in the public sector. In short, an API programming interface, is a standard that allows two different applications to communicate with each other and exchange data.

By providing the data, it is easier to provide AI models with the necessary data. APIs also offer added value when it comes to distributing and reusing functionalities. In this way, an ecosystem of useful, communicating applications can be created.

With new technological breakthroughs, the possibilities of artificial intelligence will only increase. The development of technology is an odyssey whose end nobody knows for the time being. So we can also expect some exciting things from the public sector.

This is an editorial contribution in collaboration with Smals. More information about the company’s services can be found here Here rightly so.