With the rapid growth and deployment of AI-fueled technologies, the need for “observability” within network assets is becoming an even more important necessity to maintain control over technology operations, a Government Accountability Office (GAO) official said on Dec. 12 at the Datadog Observe conference.
Taka Ariga, chief data scientist at GAO, discussed the need for observability both at his agency and across Federal agencies.
“As we navigate this new digital era, the way that we verify cybersecurity control, the way that we verify privacy considerations, the way that we verify a number of these new-era compliance kinds of issues, I think, [will] begin with observability,” said Ariga.
“If you’re not able to see something, how would you know what exactly is going on,” the GAO official asked.
During his presentation, Ariga cited an example of how GAO could practice observability in utilizing extended-reality headsets to collect data in certain situations that the agency had not been able to do in the past.
“That’s an example where I think observability becomes a paramount issue for us; we need to be able to see how data is transmitted, how cybersecurity controls are interacting with each other, and how various cloud services are being invoked as we’re doing these kinds of site visit activities,” said Ariga.
The need for observability also comes into play as Federal agencies look to adopt AI tech and in particular, large language model-based applications, he said.
“As I mentioned GAO has deployed our sort of baseline large language model, just for kicks I asked a question. ‘What was Abraham Lincoln’s opinion on GAOs AI accountability framework?’ It’s not a question that should exist … but if it exists, you should be very skeptical,” said Ariga.
“Silly example, I know, but I think this underscores if your workforce doesn’t have that level of digital literacy, or frankly, observability, you might actually take that as a fact,” he said.
“This is why it is important for us to understand the kind of traffic that is not only incoming to GAO, but also what’s leaving GAO,” Ariga said.
