Data is Ground Truth For Government AI

November 09, 2020

In the public consciousness, AI conjures Hollywood images of metaphysical life forms (Steven Spielberg’s “AI”), holograms (Obi-Wan Kenobi in “Star Wars”), and malware (HAL in “2001: A  Space Odyssey”). 

Meanwhile, back in the real world, the DNA for today’s AI in government may be more mundane, but it is no less profound or game-changing — it all comes down to data, and tons of it. 

Data is multiplying so exponentially that new metrics are being coined to keep up with the astronomical volume — beyond terabytes and petabytes, data now is being measured in zettabytes, and even, eventually, yottabytes. 

In 2020, it’s estimated 40 times more bytes of data will have been generated “than there are stars in the observable universe,” according to Jeff Desjardins of Visual Capitalist.

Data is the elemental, granular ground truth that animates AI’s catalytic role in the enterprise of government culture. 

Government agencies are becoming more creative and adventurous in their use of data, both for the betterment of society, and to improve management of internal agency resources. 

Data is being used at the Department of Transportation to upgrade the safety of surface transportation. 

The Federal Trade Commission is working with private industry to battle the onslaught of unsolicited robocalls that citizens find intrusive.

Thanks to AI and machine learning, the U.S. Small Business Administration was able to speed the processing of applications in the Covid-19 Paycheck Protection Program (PPP), while at the same time efficiently detect malign attempts by bad actors to file fraudulent claims.

The buy-in on how pivotal a role AI and data management can and should play in our lives starts at the very top. In fact, it’s a Chief Executive mandate, spelled out in a document titled “The President’s Agenda; Modernizing Government for the 21st Century.”

The document’s purpose, in its own words, is to present “a long-term vision for modernizing the Federal Government in key areas that will improve the ability of agencies to deliver mission outcomes, provide excellent service, and effectively steward taxpayer dollars on behalf of the American people.”

It notes that public trust in government is “near historic lows,” as surveyed by Pew Research Center. The percentage of Americans “who say they trust government to do what is right just about always/most of the time” has plummeted from above 70% in the late 1950s to below 20% in the 21st Century.

Clearly, as the President’s Agenda underscores, if it is to heed the vox populi by adopting practices and policies that serve the public good, the U.S. government needs to make not just incremental changes but quantum leaps. That’s where AI and data enter the conversations that are being held throughout the federal bureaucracy.

“Modernizing Federal Government,” states the Agenda, “represents a profound bipartisan opportunity to work across branches of Government … to align the mechanics of government to better meet America’s needs.” 

That rationale is the impetus for the Federal Data Strategy: “To leverage data as a strategic asset to grow the economy, increase the effectiveness of the Federal Government, facilitate oversight, and promote transparency.”

It offers examples of specific benefits attainable from a data strategy that forms a bridge to “paperless government” — improve efficiency and effectiveness of administrative services across government; develop a workforce for the 21st century; shift from low-value to high-value work; improve customer experience with federal services.

The overall goal is to improve the customer experience in dealing with federal services. Advancements in digital services are a major accelerator to make that happen, by streamlining and upgrading the customer experience, whether online, in person, or by telephony. 

Another critical role for AI-enabled data in the government’s self-improvement plan is to help “free up tens of thousands of hours that Federal personnel spend annually on meaningless compliance — time better spent pursuing mission outcomes.”

The Federal Data Strategy’s Data Incubator Project gathers and publicly shares use case examples of how data can be leveraged as a strategic asset “to achieve agency mission and public interest outcomes.” ( 

At that online location, you’ll find summaries of use cases that are user-generated. One case reads, “When USPTO (U.S. Patent and Trade Office) made data available for AI/ML models, third party partners were able to create new search tools for more efficient patent processing.”