News BriefsTechnology

National Academies of Science on: Implications of Artificial Intelligence for Cybersecurity

The National Academy of Sciences has been working for 155 years to advance our nation’s understanding of key technology issues. One of the ways they help advance knowledge is bringing together thought leaders and highly qualified experts, academics and practitioners to work on issues and then publish results for all to see. The academies have just published a recap of a March 2019 workshop held on the topic of the Implications of Artificial Intelligence for Cybersecurity.

You can order a hard copy or download a free PDF at:

https://www.nap.edu/catalog/25488/implications-of-artificial-intelligence-for-cybersecurity-proceedings-of-a-workshop

We found this report to be a good foundational overview of the status of AI and ML in support of cybersecurity. It also contained a few good sections that point to how AI and ML are being used by attackers, and a few suggestions that can inform actions by defenders.

Practitioners have seen AI and ML in cybersecurity for at least a decade, but it is good to have strategic thinkers like the ones brought together by the academies discuss and form a baseline understanding of the state fo the art so more improvement can be done.

But there is another topic needing far more research and work that was just barely touched on by the workshop. AI solutions are different than other software/data architectures. And they need more than just basic cyber defense. So we need not just conversations about how AI helps cybersecurity, but about how cybersecurity can be done for AI.

The parts of the workshop report that hit on that topic were scant and the views provided did not seem to be informed by practitioners working that topic. That is probably just a factor of how new this part of the discipline is.

For more on that, OODA network members can access our reporting at:

 

 

 

 

Bob Gourley

Bob Gourley