Artificial Intelligence (AI) is the new buzz word.  AI is seemingly everywhere! AI tools and algorithms are becoming more sophisticated. Processing power is strong, readily accessible, and cheap to use. And the world is becoming a digital marketplace with huge amounts data available. AI is poised to take-off!  This primer will give you an idea of what you can expect from AI-driven solutions as we enter a new decade.

What is AI? 
AI is the simulation of human intelligence processes by machines or computer systems. These processes include learning (the acquisition of information and rules for using that information), reasoning (using rules to reach conclusions) and self-correction.

How is AI evolving?  
Initially, AI was crunched on powerful, centralized computers. Today, you see AI being pushed to the edge—computer processing is executed on distributed nodes such as smart devices.  And those devices will increasingly know who you are, what you like, and how you do things.  The age of personalization will be driven by AI’s ability to understand how someone responds to their environment. You will also start seeing “edge decision-making” in which an “Internet of Things” device can leverage the vast experience of similar devices in different locations to make a decision instead of waiting for one to be made on its behalf. Further innovation with AI will be driven by its ability to correlate unrelated and new data sets in order to uncover relevant connections that will generate fresh and innovative improvements for our lives.
 
Why is AI for government exciting?
Governments collect enormous amounts of data. AI is driving better decision-making by quickly and intelligently processing ever-growing digital archives. Cloud architecture not only provides access to these large data volumes, it provides the raw computing power to make sense of it. Governments are turning to AI to improve operations and reduce busywork plaguing overworked staff.

For example, the SEC uses AI to help identify scammers who may have engaged in insider trading.  In the past, research analysts manually scrutinized disparate data sets for relationships weeks after transactions were completed. AI now rips through volumes of data in real time to flag possibly questionable transactions.1 

Many government departments are already using AI to streamline call centers. By making use of AI-powered chatbots, call centers are improving the customer experience, as well as improving response time and delivering more accurate information with fewer resources.2  As an example, the U.S. Army’s interactive virtual assistant, SGT STAR, fielded over a half million questions with 92% accuracy and resulted in a fourfold increase in the time users stayed in session.3 

Across all branches, government work is often content-transfer intensive. The data elements from one form often need to be transferred to another form or to a completely different system. Using AI-powered Robotic Process Automation, content can be automatically transferred between systems and forms. This, for example, allows a social worker to spend more personal time with individuals and less time on data entry.4 

 
What areas of AI create concerns?
As with any powerful technology, AI can be deployed to cause harm. One such frightening area of AI use today is the weaponization of AI outside of our borders. Our competitor nation-states are applying AI in cyber-attacks, attack drones, and other malicious projects. 

Without proper regulation, AI can be a recipe for disaster, as highlighted in a growing number of popular movies that showcase a chilling projection of a future with AI (think “Terminator”). We need to be cognizant of what we trust AI to do. On one extreme, we may consider only entrusting our AI tools to make simple decisions. On the other end, complex, life-altering decisions might require some level of human intervention. As a result, it will be important to keep control over how much we allow AI to make decisions that impact and run our government and our lives. 

The Department of Defense (DOD) is working diligently to ensure that any applications of AI adhere to strict ethical standards and longstanding DOD norms and values. Released last year, the Department’s first-ever AI strategy explicitly requires human operators to have the ability to override any decisions made by a military AI-powered system.5 

Over the last decade, our nation has permitted AI to develop unobstructed by federal regulations or standards. Critics are questioning AI’s impact on privacy and civil liberties, as well as the transparency and security of its underlying algorithms. Around the globe, scientists are drafting standards to build trust in AI algorithms that today often operate in a black box. In May 2019, the U.S. was among 42 countries to approve a new international agreement for building trustworthy AI. Adherence to the agreement will foster innovation and trust in AI as it establishes principles for the responsible development and stewardship while ensuring respect for democratic values.6 
  
What does the future of AI hold? 
Most government departments around the world are understaffed and they face massive backlogs in administration. Unemployment is at an all-time low, and resources are hard to come by, making the delivery of government services inconsistent. AI can begin to take on labor-intensive administrative tasks, speed up internal processes and increase productivity - improving the lives of citizens.


It’s already in the works. After losing more than 16% of its funding and nearly a third of its employees since 2010, the IRS is turning to AI to take over parts of its contracting process. The GSA is deploying dozens of bots within the organization to automate tasks and streamline workflow. And in May 2019, the HHS signed 57 vendors onto a $49 million contract vehicle intended to deliver AI tools.  Beyond administrative relief, federal law enforcement is turning to more advanced AI tools, namely facial recognition, to investigate crimes and to control our borders.7 


At defense and intelligence agencies, officials are leveraging AI to make sense of the huge volumes of information flooding in from across the globe. As the digital economy opens up more data to more people than ever before, national security could hinge on how effectively we use AI to find the signal(s) in the noise.


Conclusion


Heading into the next decade, AI-powered solutions hold the promise of addressing some of our most critical challenges. Our digital future and AI are intertwined.  Within government, I believe it is fair to expect to see a substantial increase in the use of AI to help drive faster and more informed decision-making.

Footnotes