Page 183 - Cyber Defense eMagazine July 2024

But with emerging technology like artificial intelligence, new standards like this are just beginning to take
shape.

Numbers have shown the federal government has an appetite for AI. According to a report from Stanford
University, U.S. Defense and Federal Civilian agencies spent nearly $3B on AI solutions. This illustrates
that the federal government recognizes the benefits and needs to adopt artificial intelligence to remain
competitive and protect our national security. But how do technology companies become mission-ready
for these needs?

The Intersection of Standards

There are a few recent mandates around the federal use of AI such as the Office of Management and
Budget’s newly released Memo M-24-10. This states government agencies must meet and implement
mandatory AI safeguards that provide more reliability testing, transparency and testing of AI systems.
Agencies must meet these standards by December 1, 2024.

This is where it gets complicated. Since many commercial AI solutions are delivered using cloud services,
these AI solutions must be FedRAMP accredited.

With the rapid adoption of AI, there are now federal agency-specific use cases that detail the intersection
of AI and cloud services. For example, the Department of Labor (DOL) has several projects utilizing cloud
based commercial off the shelf NLP models for language translation, claims document processing and
website chatbots. The United States Treasury has similar use cases.

These use cases, with both cloud and AI integration, are subject to FedRAMP compliance already.

Meeting New Benchmarks and Beyond

Regardless of whether a technology company is providing a cloud-based AI service or just a typical AI
model, there are a few steps that can be taken now to accelerate the use of AI by building upon existing
frameworks like FedRAMP.

Compliance can be achieved at a faster pace with an authority-to-operate (ATO) system to create an
overlay for AI that is based on NIST AI RMF and NIST SP 800-53. By applying an ATO to AI, agencies
can tailor, extend and augment existing guidelines and accelerate the integration of AI systems and
safeguards.

Another helpful resource comes from the FedRAMP Program Management Office which recently
published the Emerging Technology Prioritization Framework, designed to accelerate the availability of
FedRAMP accredited Gen AI cloud solutions for federal agencies.

To jumpstart the availability of AI solutions, the FedRAMP PMO published a draft prioritization framework
that defines the initial categories of Generative AI solutions and the benchmarks that will be used to drive

Cyber Defense eMagazine – July 2024 Edition 183
Copyright © 2024, Cyber Defense Magazine. All rights reserved worldwide.