Software Engineering Institute (SEI) Podcast Series

Available Episodes

5 of 421

From Data to Performance: Understanding and Improving Your AI Model
Modern data analytic methods and tools—including artificial intelligence (AI) and machine learning (ML) classifiers—are revolutionizing prediction capabilities and automation through their capacity to analyze and classify data. To produce such results, these methods depend on correlations. However, an overreliance on correlations can lead to prediction bias and reduced confidence in AI outputs. Drift in data and concept, evolving edge cases, and emerging phenomena can undermine the correlations that AI classifiers rely on. As the U.S. government increases its use of AI classifiers and predictors, these issues multiply (or use increase again). Subsequently, users may grow to distrust results. To address inaccurate erroneous correlations and predictions, we need new methods for ongoing testing and evaluation of AI and ML accuracy. In this podcast from the Carnegie Mellon University Software Engineering Institute (SEI), Nicholas Testa, a senior data scientist in the SEI's Software Solutions Division (SSD), and Crisanne Nolan, and Agile transformation engineer, also in SSD, sit down with Linda Parker Gates, Principal Investigator for this research and initiative lead for Software Acquisition Pathways at the SEI, to discuss the AI Robustness (AIR) tool, which allows users to gauge AI and ML classifier performance with data-based confidence.
--------
26:42
--------
26:42
What Could Possibly Go Wrong? Safety Analysis for AI Systems
How can you ever know whether an LLM is safe to use? Even self-hosted LLM systems are vulnerable to adversarial prompts left on the internet and waiting to be found by system search engines. These attacks and others exploit the complexity of even seemingly secure AI systems. In our latest podcast from the Carnegie Mellon University Software Engineering Institute (SEI), David Schulker and Matthew Walsh, both senior data scientists in the SEI's CERT Division, sit down with Thomas Scanlon, lead of the CERT Data Science Technical Program, to discuss their work on System Theoretic Process Analysis, or STPA, a hazard-analysis technique uniquely suitable for dealing with AI complexity when assuring AI systems.
--------
36:14
--------
36:14
Getting Your Software Supply Chain In Tune with SBOM Harmonization
Software bills of materials or SBOMs are critical to software security and supply chain risk management. Ideally, regardless of the SBOM tool, the output should be consistent for a given piece of software. But that is not always the case. The divergence of results can undermine confidence in software quality and security. In our latest podcast from the Carnegie Mellon University Software Engineering Institute (SEI), Jessie Jamieson, a senior cyber risk engineer in the SEI's CERT Division, sits down with Matt technical director of Risk and Resilience in CERT, to talk about how to achieve more accuracy in SBOMs and present and future SEI research on this front.
--------
23:14
--------
23:14
API Security: An Emerging Concern in Zero Trust Implementations
Application programing interfaces, more commonly known as APIs, are the engines behind the majority of internet traffic. The pervasive and public nature of APIs have increased the attack surface of the systems and applications they are used in. In this podcast from the Carnegie Mellon University Software Engineering Institute (SEI), McKinley Sconiers-Hasan, a solutions engineer in the SEI's CERT Division, sits down with Tim Morrow, Situational Awareness Technical Manager, also with the CERT Division, to discuss emerging API security issues and the application of zero-trust architecture in securing those systems and applications.
--------
17:41
--------
17:41
Delivering Next-Generation AI Capabilities
Artificial intelligence (AI) is a transformational technology, but it has limitations in challenging operational settings. Researchers in the AI Division of the Carnegie Mellon University Software Engineering Institute (SEI) work to deliver reliable and secure AI capabilities to warfighters in mission-critical environments. In our latest podcast, Matt Gaston, director of the SEI's AI Division, sits down with Matt Butkovic, technical director of the SEI CERT Division's Cyber Risk and Resilience program, to discuss the SEI's ongoing and future work in AI, including test and evaluation, the importance of gaining hands-on experience with AI systems, and why government needs to continue partnering with industry to spur innovation in national defense.
--------
30:18
--------
30:18