[ad_1]
As a part of an ongoing effort to maintain you knowledgeable about our newest work, this weblog publish summarizes some not too long ago revealed SEI studies, podcasts, and webcasts highlighting our work in safety incident response, DevSecOps, and cybersecurity coaching. Now we have additionally included a collection of three white papers that define the three pillars of AI engineering: scalable, strong and safe, and human-centered.
These publications spotlight the newest work of SEI technologists in these areas. This publish features a itemizing of every publication, writer(s), and hyperlinks the place they are often accessed on the SEI web site.
Human-Centered AI
by Contributor Hollen Barmer, Rachel Dzombak, Matt Gaston, Jay Palat, Frank Redner, and Carol J. Smith
Human-centered AI methods are designed to work with, and for, individuals. As the need to make use of AI methods grows, human-centered engineering ideas are crucial to information system growth towards efficient implementation and reduce unintended penalties. We determine three particular areas of focus to advance human-centered AI:
- the necessity for designers and methods to grasp context of use and sense modifications over time
- growth of instruments, processes, and practices to scope and facilitate human-machine teaming
- strategies, mechanisms, and mindsets to have interaction in crucial oversight
For every space, we determine ongoing work as effectively and challenges and alternatives in creating and deploying AI methods with confidence.
Obtain the white paper.
Strong and Safe AI
by Contributor Hollen Barmer, Rachel Dzombak, Matt Gaston, Eric Heim, Jay Palat, Frank Redner, Nathan M. VanHoudnos
Strong and safe AI methods reliably function at anticipated ranges of efficiency, even when confronted with uncertainty and within the presence of hazard or risk. These methods have built-in constructions, mechanisms, or mitigations to stop, keep away from, or present resilience to risks from a specific risk mannequin. We determine three particular areas of focus to advance strong and safe AI for protection and nationwide safety:
- bettering the robustness of AI parts and methods
- designing for safety challenges in trendy AI methods
- creating processes and instruments for testing, evaluating, and analyzing AI methods
For every space, we determine ongoing work in addition to challenges and alternatives in creating and deploying AI methods with confidence.
Obtain the white paper.
Scalable AI
by Contributor Hollen Barmer, Rachel Dzombak, Matt Gaston, Jay Palat, Frank Redner, John Wohlbier
Scalable AI is the flexibility of AI algorithms, information, fashions, and infrastructure to function on the dimension, pace, and complexity required for the mission. Scalability is a crucial idea in lots of engineering disciplines and is essential to realizing operational AI capabilities. We determine three areas of focus to advance scalable AI:
- scalable administration of knowledge and fashions to beat information shortage and assortment challenges
- enterprise scalability of AI growth and deployment
- scalable algorithms and infrastructure
For every space, we determine ongoing work in addition to challenges and alternatives in creating and deploying AI methods at scale.
Obtain the white paper.
The Sector CSIRT Framework: Creating Sector-Based mostly Incident Response Capabilities
by Justin Novak, Brittany Manley, David McIntire, Sharon Mudd, Angel Luis Hueca, Tracy Payments
The U.S. Division of State, Workplace of the Coordinator for Cyber Points commissioned the Software program Engineering Institute (SEI) to create the Sector CSIRT Framework for (1) creating a sector-based laptop safety incident response and coordination functionality and (2) integrating this functionality into a bigger nationwide cybersecurity ecosystem as relevant. The framework is a information for serving to events develop the insurance policies, processes, and procedures essential to operationalize a sector laptop safety incident response staff (CSIRT), a uniquely tailored, specialised incident response staff. The framework outlines a course of that strikes the sector CSIRT from idea to actuality. The framework helps the staff creating the sector CSIRT perceive the present situations of incident response within the sector (i.e., the as-is state) and tips on how to transfer it to a sturdy working state (i.e., the to-be state). It bridges the hole between these two states utilizing a well-defined roadmap and implementation course of.
The Sector CSIRT Framework is meant for people and organizations—together with CSIRT managers, nationwide CSIRTs, and others—who’re creating or implementing a sector CSIRT. Utilizing this framework, these people or organizations can transfer a sector CSIRT from an idea to the fact of a totally operational staff.
by Rachel Dzombak, Jay Palat
On this SEI Podcast, Rachel Dzombak and Jay Palat, each researchers with the SEI’s Rising Know-how Middle, focus on development within the area of synthetic intelligence (AI) and the way organizations can rent and prepare employees to benefit from the alternatives afforded by AI and machine studying—and the crucial want for an AI engineering self-discipline to develop the AI workforce.
Obtain/view/take heed to the podcast.
Is Your Group Prepared for AI?
by Rachel Dzombak, Carol Smith
On this SEI Podcast, digital transformation lead Dr. Rachel Dzombak and analysis scientist Carol Smith, each with the SEI’s Rising Know-how Middle at Carnegie Mellon College, focus on how AI Engineering can help organizations to implement AI methods. The dialog covers the steps that organizations must take (in addition to the exhausting conversations that must happen) earlier than they’re AI prepared.
Obtain/view/take heed to the podcast.
Software program Provide Chain Issues for DevSecOps Applications
by Aaron Okay. Reffett and Richard Laughlin
In a DevSecOps world the software program provide chain extends past libraries upon which developed software program relies upon. On this webinar we take a look at the SolarWinds incident as a worst-case exemplifying the breadth of the software program provide chain points confronting advanced DevSecOps packages. We are going to discover the vital architectural features of DevSecOps which can be impacted by the software program provide chain that require consideration and potential mitigations to detect and reply to potential incidents.
What attendees will study:
- The software program provide chain challenge is broad and impacts a number of features of DevSecOps.
- Applications want to pay attention to how the software program they leverage presents dangers.
- Mitigation methods have to be put in place to deal with potential points on the architectural stage.
How Do We Train Cybersecurity?
by Rotem D. Guttman and Josh Hammerstein
How do you educate cybersecurity to a center faculty pupil? To a soldier? To a few of the greatest hackers within the nation? How do you consider all of those audiences’ expertise? Cybersecurity coaching has been an ongoing problem for many years. The important thing to creating the most effective use of your coaching greenback is to craft coaching that matches your viewers’s wants and engages them in a significant method. If you create an expertise so enthralling that your viewers is logging in on nights and weekends simply to proceed taking part, the worth of immersive coaching actually shines. Be a part of us throughout this webinar as Rotem Guttman shares the teachings he’s realized over a decade of creating participating, immersive coaching and analysis environments for quite a lot of audiences.
What attendees will study
- tips on how to make cybersecurity coaching participating
- what motivates several types of learners
- the historical past of enhanced cybersecurity coaching on the SEI
[ad_2]
