AI Boosts Security and Operational Capability

Artificial Intelligence (AI) is emerging as a potentially potent ingredient for cyber criminals’ exploits and malware attacks. On the other hand, AI is also becoming key to defending against those new threats and expanding agency capabilities.

Emerging natural language AI applications, such as ChatGPT-4, leverage artificial intelligence, deep learning, and big data to generate all manner of apps, articles, and other information-based content, as well as writing and debugging computer coding, without human intervention.

Hackers are figuring out how to leverage all those capabilities.

AI, deployed comprehensively and coherently across federal agency enterprises, can support the next-step data analytics capabilities, and foster more secure algorithms that can become integral pieces of strong cybersecurity plans, as well as more human-oriented operations.

Secured Data Is AI Key

“The cybersecurity threat from ChatGPT is just emerging, as are a host of AI-driven enterprise applications”, said Jon S. Kim, SVP of Solutions and Services at Presidio Federal, a provider of secure mission-critical custom IT products and services to the federal government. Unsecured data sources could spur federal agencies to move more securely along on their AI paths, according to Chris Maestas, Chief Executive Architect at IBM, a pioneer of AI technologies and services. IBM’s reach with AI-powered services and capabilities is long and wide, providing AI-enhanced storage, processing, and process automation.

Today, Maestas notes, the triage of incoming data has become a crucial part of using big data to run applications and secure operations. “Doing an initial scan of incoming data can reveal something that could wake up and surprise us,” said Maestas.

AI, with its digital processes including machine learning and natural language processing, can help agency security teams procure threat intelligence from millions of sources and sort through that intelligence quickly and efficiently.

According to Maestas, efficient processing sets secure data foundations that applications draw from. Data at agencies is being generated from all manner of sources, from drones to high-speed cameras and a myriad of other devices. The data captured in those devices is placed in transient storage and tagged, then passed on to AI and other systems that use big data. Drawing data from archives can also have some security issues if it’s used to power an AI system. AI can help ensure that data is secured before it is drawn into applications.

Operationalize AI for Security

Using AI to look at incoming data consistently and automatically over long periods can turn up potential data irregularities and help secure IT operations, according to Maestas and Kim. “In security, AI can look at very long data baselines for even very minor changes, that humans can’t do,” said Kim. “AI can look for things that might go unnoticed in the big data lakes that cybersecurity needs to work from.”

That kind of application of AI capabilities is sometimes referred to as AIOps. AIOps can sharpen data, data analysis and reporting. AI can help aggregate oceans of data generated by IT infrastructure components, applications, performance monitoring tools and service ticketing systems. 

“AIOps can help personnel deal with the increasing volumes of data security, including false positives” that analysis turns up, said Maestas.

AI can also help diagnose root causes of problems and automatically report them to IT and DevOps for rapid response and remediation, and even automatically resolve them without human intervention. 

“Ideally,” said Maestas, “AI could help cybersecurity departments do ‘on the fly’ patching that would immediately provide protection from quickly evolving cyberthreats in an increasingly dynamic cybersecurity environment.”

Improve Everyday Processes and Services

“There are less dramatic areas where AI can help make major improvements,” said Kim. IT help desk functions and even employee training programs can leverage AI.

For instance, virtual help desk assistants using the natural language comprehension and processing capabilities inherent in AI, can tackle everyday IT questions and duties for employees, allowing them to get to other, more important tasks. “Help desks powered by AI”, said Kim, “can help sort through support tickets and even suggest or implement solutions.”

Enterprise network engineers can also tap into it to perform everyday tasks such as monitoring, support, as well as everyday manual processes, he said.

Comprehensive Platform Solutions Solve Problems

A comprehensive global data platform that harnesses AI can help make sense of data coming into enterprises and agencies, as well as secure it, according to Maestas. For example, AI incorporated into IBM’s Storage Scale platform powers its data handling capabilities for high-performance and next-generation data services. IBM Storage Scale’s high performance unstructured data management solution can connect endpoint technologies, audit incoming data and encrypt it. It can then be coupled with a data cataloging service that can do metadata tagging and deep inspection in real time, which can then be turned over to any other analytics that are necessary, according to Maestas.

This content is made possible by our sponsor. The editorial staff was not involved in its preparation.

NEXT STORY: Advancing Zero Trust Readiness: What Federal Agencies Can Do to Move the Needle on Internal Practices