AI in Government – Current AI Projects in the Public Sector

Government interest in AI has picked up recently, and lots of government officials are actually beginning to question the very same issues company professionals have been asking 2 or maybe 3 years back. Governments as well as big NGOs are actually beginning to purchase AI, spending budget as well as time on pilot programs for different AI talks and applications with folks of the field on the future implications of the technology.

With the summer, our founder spoke at a joint UN INTERPOL conference of Singapore on the implications of AI for police. Much more recently, he spoke at the UN in Geneva on the present uses of AI, completed a speaking tour for the World Bank across Asia and Europe during which he discussed how AI might help improve healthcare in developing nations, as well as spoke with generals as well as colonels learning at National Defense Faculty in Washington DC.

As machine learning abilities develop, world governments will probably spend a lot more in AI studies. IDC estimated the US government’s investment in artificial and cognitive intelligence technologies will develop at a CAGR of 54.3 % between 2018 2021. Presently, local and national governments work with academia as well as business leaders to bring to life AI tasks which range in software from coordinating visitors to digitizing backlogs of government documents.

We researched the public sector to find out exactly where AI was driving worth for governments of all sizes and to get a concept for what government officials are actually contemplating the technology.

This particular report covers AI tasks which governments have piloted, implemented, or perhaps planned throughout 3 applications:

  • Citizen Engagement
  • Improving Operational Efficiency
  • AI Strategies for Policy and Legislation

Which said, AI uses in government are nascent nowadays. Governments might have to conquer a number of hurdles to effectively adopt the technology:

  • Government executives could possibly have to invest in upgrading the legacy of theirs IT methods. Not doing so risks compatibility problems when the time comes to incorporate the program.
  • Government leaders really should make certain the departments of theirs have the essential computing materials for an AI task. Cloud computing remedies might suffice for sure AI programs, but much more weighty information crunching could call for costlier graphical processing units.
  • Government leaders must be well prepared to employ subject matter as well as information scientists experts that could collaborate to figure out the low hanging fruit issues that AI might possibly solve for a certain government division.

Below, we detail several of the present AI tasks which governments have worked on. One typical theme with these tasks appears to be a large number of governments are actually engaging in exploratory AI tasks through collaborative public private partnerships with business as well as academia.

Citizen Engagement

New York City Department of Social Services

The brand new York City Department of Social Services (DSS) has a staff of about 18,000 and also provides public assistance as well as employment services to more than 3 million citizens. In among the initiatives of theirs, known as the Supplemental Nutrition Assistance Program (SNAP), workers of the federal government organization required to process roughly 70,000 SNAP uses a month.

Citizens wishing to avail the DSS services had been mostly walking into the government’s brick as well as mortar service facilities to post the SNAP uses of theirs. With such a lot of uses and limited hand online resources, DSS determined to digitize as well as automate the task by providing a self service on-line portal known as the ACCESS HRA (AHRA).

The best objective of the project was improving the internet experience for site visitors and minimize the amount of individuals entering the brick of theirs as well as mortar service facilities. The AHRA portal purportedly allows users inquiring about disability benefits, income assistance, food assistance, and health insurance to view the case status of theirs, account balances, as well as e notices.

DSS and IBM interested in a private public partnership to build the portal. The IBM Watson Health Government staff was tasked with finding a scalable resolution to optimize outreach to the client base. IBM was permitted to access information from PC banks at fifteen of DSS’ bodily service facilities. It allowed IBM to access user feedback which illuminated trends in what folks desired to find out through the system.

The partnership started a multidisciplinary team with a mix of private and public sector business as well as tech professionals. The staff worked out of the DSS design studio, in which components of the self service portal had been created jointly.

According to Lauren Aaronson, Assistant Deputy Commissioner for the Office of Business Process Innovation at DSS, after the task with IBM, more than seventy five % of SNAP applications had been submitted online, and foot traffic in the physical service facilities was decreased by thirty %.

The NYC DSS wasn’t explicit in the way they leveraged IBM Watson’s machine learning abilities, but Watson is actually predicated on AI. As a result, we might be in a position to infer the product’s use. Watson might have allowed the DSS to get a digital portal with search abilities built on natural language processing. We detail a use case because of this capability in banking in the report of ours on natural language processing in banking.

Furthermore, IBM Watson might have allowed NYC DSS to digitize their newspaper files using machine vision. For doing so, the DSS would theoretically be in a position to browse the document resulting scans of theirs for different keywords related to casework as in case they were more structured papers which had been entered digitally to start with.

US Citizenship and Immigration Services

Throughout 2015, the US Immigration and Citizenship Services (USCIS) announced the launch of a chatbot named Emma. Emma can reportedly answer questions on immigration and get visitors to the appropriate page of the USCIS website. The USCIS was getting a high volume of client queries, with more than fourteen million calls on immigration issues every year.

Users are able to click the “Ask a Question” and  “Need Help? Question Emma” links in top of the right corner of the USCIS webpage. In cases in which Emma can’t search for the proper info, the software program automatically redirects the consumer to a man agent.

We couldn’t find some strong case research reporting good results with the chatbot. We are able to, nonetheless, infer that the chatbot utilizes natural language processing as well as natural language generation to the office. The algorithm behind the chatbot was apt given a huge number of text messages relating to a variety of typical immigration concerns as well as requests for services , like requests for the condition of one’s visa applications.

These emails will have been labeled by humans as different kinds of inquiries, for example “request for an application form,” or even a few other. This might have educated the algorithm to know the chains of written text which make up a message involving one of those sorts of inquiries when provided with a message that wasn’t labeled. The chatbots would then make a reaction to the user.

Infocomm Development Authority of Singapore

The federal government of Singapore recently announced a partnership with Microsoft to create conversational interfaces for their Smart Nation Initiative. Vivian Balakrishnan, Minister for Foreign Affairs as well as head of the Smart Nation Initiative, announced the government’s plans to check out the usage of chatbots for selected public services. To this conclusion, the Infocomm Development Authority of Singapore (IDA) is actually working alongside Microsoft to create conversational intelligence into present public services as a proof of concept task. The aim for the task is making online public service sites much more accessible and engaging for those citizens.

The IDA had previous experience creating a virtual assistant called “Ask Jamie” which may be applied on the sites of other governmental agencies to answer questions in particular domains. When users go to the agency site, a chat window instantly opens up, and Jamie is able to react to user inquiries by means of natural language generation.

The IDA claims that Ask Jamie may additionally be taught to question follow up questions if the search query from a user is simply too wide. In case the issue isn’t a thing for which Jamie might find a solution, the chatbot purportedly transfers the user to a man representative in the federal government organization.

As for the Smart Nation Initiative, Microsoft is designed to finish the proof of concept in 3 steps:

  1. The very first model chatbots are going to aim to browse and retrieval of info from existing databases making use of natural language processing. For instance, an end user may be in a position to search for the proper web page for filing a visa form within a government agency’s site.
  2. The next model chatbots would have extra features that allow users to complete basic tasks, like booking a meeting or even creating a transaction on a government site. These chatbots could be in a position to book a scheduled appointment with a consulate and enable users to spend some fee active in the progression.
  3. The 3rd generation chatbots are actually aimed to be personalized, in which people are able to ask questions and receive highly personalized info. For instance, the end user may get info about what files they may need to post to the agency based on existing data for that specific user.

Based on the press release, Microsoft will work with the federal government of Singapore to open a Centre of Excellence (CoE) for Leadership in Digital Transformation for Industry as well as Government. The CoE will even help support the evidence of concept with IDA to produce abilities which can fulfill multilingual populations as well as individuals of different ages.

Improving Operational Efficiency

Automated Traffic Control

City of Pittsburgh

The City of Pittsburgh collaborated with Rapid Flow Technologies to develop SURTRAC(Scalable Urban Traffic Control), an automated traffic seo as well as management software. Community traffic management departments are able to make use of SURTRAC to control traffic flows through a number of intersections and make use of AI to enhance the visitors devices toward decreased travel times, decreased number of traffic stops, as well as decreased wait times.

Pittsburgh Department of Transportation worked with Rapid Flow Technologies in 2012 to set up the ca at a pilot project in the East Liberty neighborhood of Pittsburgh. Surtrac was linked to a network of 9 traffic signals in 3 main roads in Pittsburgh (Penn Circle, Penn Avenue, and Highland Avenue).

Surtrac received information from receptors or maybe cameras embedded in the website traffic signals and calculated the traffic allocation projections of the selection of automobiles flowing out at each intersection. The program also re calculated this projection every 1 2 secs constantly to allow it to be highly reactive to unexpected changes in traffic situations.

We are able to infer that it was made possible because of to printer vision technologies. The machine learning algorithm behind the software program was apt fed labeled pictures of automobiles at different angles, in different lightnings, going at different speeds, and also in different volumes of traffic. This might have educated the algorithm to discern traffic volume when provided with traffic footage which wasn’t labeled.

Fast Flow claims SURTRAC helped the city reduce traveling time by twenty five %, traffic stops by thirty %, wait time by forty %, and total emissions by twenty one % over the course of the pilot. Fast Flow then collaborated with neighborhood Pittsburg administration to grow the answer to various other areas of the city, bringing the total of SURTRAC installed traffic signals to about fifty.

Greg Barlow is CTo at Rapid Flow Technologies. He holds a PhD in Robotics from Carnegie Mellon Faculty.

Predicting Fire Risk in Buildings

Atlanta Fire Rescue Department

AI researchers from Georgia Institute of The University, Emory University, and Technology of California, Irvine worked with the Atlanta Fire Rescue Department (AFRD) to build a predictive analytics software targeted at identifying buildings that have a better probability of fire incidents.

The scientists developed Firebird, which purportedly makes use of historical details for fifty eight variables made available to them by the AFRD. This includes information including property area, fire incidence, creating size, building framework, and season created for the period 2010 2014. The information was given to an AI predictive analytics software program and also the algorithms had been tweaked to forecast fire danger scores for 5,000 buildings. Based on the faculty, the program correctly predicted seventy three % of fire incidents in the structure.

AI Strategy for Policy and Legislation

The proliferation of AI in government could warrant the have for a macro level strategy for world governments, just as a result of disruptive potential of AI. Governments have to make certain that the policies of theirs as well as legislation are actually coherent with just how AI may evolve in the following ten years. Down below we provide snapshots of several of the broader initiatives by earth governments detailing the methods of theirs. These initiatives are robust and nascent relatively info on them is limited. 


Germany’s Federal Government announced that it was drafting a Strategy on Artificial Intelligence (AI) that will be presented at the 2018 Digital Summit in Nuremberg in December.

From the press release of theirs, the German Federal Government stated they are going to consult a number of pros on the application option room as well as regulations which may be required. These professionals had been from business associals as well as academia all over Germany as well as the insights of theirs will serve as a foundation for the improvement of the technique. This particular board of consultants will additionally aid with defining fields as well as objectives of action for the Strategy that will ultimately be used by the German Federal Cabinet.


National Institute of Advanced Industrial Science and Technology (AIST) and METI Ministry of Economy, Trade and Industry have announced a collaboration aimed at attracting top AI researchers for the development and commercialization of Japanese AI technologies. To this conclusion, the Japan Meteorological Agency, Japan Patent Office, Electricity, Transport and Traffic authorities are presently undertaking AI tasks which will help augment the skills of human operations in projects like patent as well as trademark examination, the control as well as forecast of traffic, as well as the improvement of predictive abilities for emergency responses


Inside our AI contained India research report, we recognized the pros we interviewed seemed to agree that having the proper AI policy in place with a mission statement as to just how AI must evolve in the nation may be needed to enhance understanding among citizens. In June 2018, the federal government think tank, National Institution for Transforming India (NITI) Aayog (Hindi for Policy Commission), provided their National Strategy for Artificial Intelligence. Furthermore, Indian government’s Commerce and Industry Department also launched a Task Force on Artificial Intelligence for India’s Economic Transformation in 2017.

There appears to be a bit of proof that Indian state governments are today using AI for particular uses including citizen engagement. For instance, The Andhra Pradesh government announced a partnership with Microsoft to have the Kaizala app. Kaizala is actually being used to gather citizen comments from social networking as well as the government’s previously existing web portals. The customer feedback is instantly sorted as well as routed to the correct department for account.

Source: published by Raghav Bharadwaj,

Anikesh is one of the brilliant editors we have in Analytics Jobs. Anikesh is a technology journalist and a techopedia who likes to communicate the latest trends around cutting-edge technologies in a way that is straightforward to assimilate.


  1. REO Professionals

    In accordance with my observation, after a foreclosed home is sold at an auction, it is common for the borrower in order to still have the remaining unpaid debt on the mortgage loan. There are many financial institutions who aim to have all charges and liens repaid by the future buyer. Having said that, depending on specific programs, laws, and state guidelines there may be several loans that aren’t easily fixed through the exchange of lending options. Therefore, the obligation still falls on the customer that has had his or her property foreclosed on. Thanks for sharing your thinking on this weblog.

  2. Virtual Assistants

    Great post. There are many reasons why delegation is crucial, especially if you are a business executive. You need to be able to focus on the tasks that create income and not get stuck by those that do not.


Leave your thoughts

Get your daily dose of analytics jobs alert, news, blogs, happenings and keep yourself updated.