As in private industry, data is the key to everything in government—creating actionable insights, rooting out fraud and waste, uncovering cyberactivity, and generally improving efficiency and productivity. It’s so important that the Biden administration last year unveiled the Federal Data Strategy Action Plan, which lays out a plan for how federal agencies will accelerate the use of data to deliver on missions, serve the public and protect resources while ensuring security, privacy and confidentiality. Individual agencies have taken note; most of them today now have Chief Data Officers, who are dedicated to finding ways to leverage the power of data to enable data-driven decision-making.

Here are four trends that will impact the government’s ability to manage their data effectively and in real time:

Government data will continue growing exponentially, making data management more challenging than ever

In most agencies, data is growing so fast, and stored in so many different, often unconnected repositories that it’s hard to keep up. With so much data coming from so many sources, agencies are looking for ways to improve the ability to ingest, analyze, and glean value from changing data quickly—ideally, in real time. Think about military operations overseas; units must be able to receive and process data in real time at the edge to make fast, accurate decisions.

One of the main sources of data growth is the proliferation of sensors and other Internet of Things (IoT) devices throughout agencies. IoT is most often used to control or monitor equipment or systems, control access to devices and facilities, track physical assets and monitor movement at borders. A GAO survey found that most federal agencies today are using IoT technologies. If the growth follows the trajectory of other industries, it will continue growing exponentially; IoT growth is expected to exceed 26% through 2029.

Increasingly, much of this data growth is being generated at the edge—near the battlefield, in vehicles, at disaster sites, processing centers and secure facilities, among others. According to IDC, data creation at the edge will grow 33 percent by 2025, accounting for 22% of all digital data being created, captured and replicated.

As if the challenge of explosive data growth in multiple places wasn’t enough, much of the data being created today is unstructured—in non-standard and disparate forms like audio and video files, sensor data, emails, images, surveillance and geospatial data and text files. Managing unstructured data and using it for analysis is challenging, but it’s important—the data often holds valuable and actionable insights.

AI and machine learning are increasingly important, and they rely on usable data

According to Gartner, 36% of government respondents in a recent survey said they planned to increase investment in artificial intelligence (AI) and machine learning. If anything, that estimate is probably on the low side. AI and machine learning are integral to gleaning insights that can assess risks, identify anomalous behavior and increase efficiencies. It helps improve financial management, fraud and cyberthreat detection, analyze personnel performance and much more.

Federal leaders also are on board; the Biden Administration launched an AI task force last year to improve access to AI. At the same time, one survey found that while agencies have all of the data they need to conduct effective analysis, it wasn’t clear how fully or effectively agencies were actually using those capabilities.

But AI and machine learning tools are only as good as the real-time data they have to work with. The key is combining AI and ML capabilities with a data platform that can ingest large amounts of data in real-time from a variety of platforms.

Data security is non-negotiable

The Federal Zero Trust Strategy requires, among other things, that security and data teams work together to improve data security. This includes creating rules to automatically detect and block unauthorized access to sensitive information. It puts great emphasis on strong enterprise identity and access controls, including multi-factor authentication. Other increasingly important data security technologies include encryption, role-based permissions and comprehensive audit trails.

There are good reasons why federal leaders can’t compromise on data security; 61% of government leaders say their organizations’ top challenge is protecting data, especially in light of the uptick of remote access. Another survey found that 62% of public sector IT and security professionals say data protection, data recovery capabilities and privileged access management are increasingly important cybersecurity measures.

Data will become even more dynamic

Data will change faster and more frequently than ever before. It will no longer be acceptable to analyze massive amounts of static data once per month, once per week, or even once per day. That means that agencies will have to glean insights from current data in real time to find new patterns and discover and act on them. Adapting to a constantly changing environment isn’t easy, but it’s critical in today’s world. It’s an important point for agencies; 93% of federal Chief Data Officers say leadership relies on data-driven insights some or all of the time.

Corralling and leveraging frequently changing data requires the ability to ingest and act on streaming data and the ability to combine that data with systems of record, third-party sources or data lakes in real time.

More than ever, data is the basis for every actionable decision federal agencies make. By understanding these important trends, agencies can make better decisions for how to steward their data moving forward.