AI is gaining momentum in the government market as IT shops look to it as a critical tool in facing cybersecurity threats.
For a long time, the term “artificial intelligence” was a sci-fi concept perceived as nothing more than Isaac Asimov stories and farfetched robotic advances.
But with stronger processing speeds and more in-depth analytics, AI has been making headway in the IT sector for system operations, cloud computing and cybersecurity. Spending on AI and machine learning, which helps make AI possible, will grow from $12 billion in 2017 to $57.6 billion by 2021, according to IDC. And it’s starting to get the attention of federal and state and local government IT personnel who see it as a way to increase and optimize automation for enhanced judgment and cost reduction.
The largest opportunity for AI is cybersecurity. Government agencies spend significant resources and people hours adapting to cyber threats while hacker technology becomes even more persistent and evolving. This is the wild west with cybersecurity and the trick is to stay one step ahead of malware, spyware and viruses that aim to corrupt and compromise sensitive processes and data. One area of concern is the Internet of Things, where AI integration can improve security measures to prevent infiltration at the many nodes that comprise the large node and sensor-heavy edge. The benefit would be that people wouldn’t have to reprogram its many layers.
Federal IT managers see cybersecurity as the largest growth area for AI across the government. But not quite yet. In the first quarter of fiscal 2018, only 21 percent of leaders surveyed by Meritalk are confident the technology is mature enough to be considered safe. Only 46 percent want the technology adopted in current form.
This isn’t surprising as the technology is still rather new in its application to government machines. And like most new IT, it’s met with a healthy dose of skepticism at first. But the number of influencers is likely to rise as success stories emerge. On the other hand, 79 percent of those polled by Meritalk agree that AI and machine learning have a place within government if done correctly.
The Department of Defense and intelligence agencies are significantly further along in adopting the technology than their civilian counterparts—over 12 percent more stakeholders are discussing it in their CIO offices. Chris Lynch, director of the Defense Digital Service, a technology shop within the White House, recently said he would like to see machine learning incorporated into the massive Joint Enterprise Defense Infrastructure (JEDI) cloud, which would make it revolutionary and set it apart as a key differentiator.
There’s also activity on the civilian side. Groups like the Department of Homeland Security’s Science and Technology Directorate and the National Biosurveillance Integration Center have begun early pilots of machine learning to spot disease outbreaks. However, it’s a bit more like advanced predictive analytics and automation than true machine learning, which raises the issue of when does a system become AI and what are the levels of maturity. Some of this has been answered for robotics but government entities will define them differently until policy is written and public nomenclature is formed.
State and local governments are also taking a close look at AI and machine learning. Stu Davis, Ohio’s chief information officer, said in October that the state will be looking at data analytics and machine learning language to be incorporated into requests for proposals. The state has machine learning pilot programs underway for primitive helpdesk functions but Davis said he’d like the IT industry to provide more guidance in how the state should be utilizing machine learning. His agencies are struggling to find areas where machine learning would help.
Simply put, AI is not just a buzzword, it’s an up and coming technology that’s finding its way into government. While we don’t yet know how much the government AI market is worth, the private sector has a tremendous opportunity that’s in its infancy. Now is the time for companies to start targeting DOD and intel agencies with AI and machine learning solutions that address cybersecurity challenges. Those early success stories can then help bring the technology into civilian agencies.