Virtual Library

Our virtual library is an online repository of all of the reports, papers, and briefings that IST has produced, as well as works that have influenced our thinking.

Submit your Content

Reports

Nuclear Command, Control, and Communications (NC3): A Primer on Strategic Warning, Decision Support, and Adaptive Targeting Subsystems 

Alice Saltini, Sylvia Mishra, Philip Reiner

viewpdf

Reports

Securing the Signal: Mitigation Strategies to Strengthen Crisis Communication Channels

Christian Steins

viewpdf

Reports

Strengthening Nuclear Crisis Communications: Steps to Implement Mesh Networks to Enhance Resilience & Security

Christian Steins

viewpdf

Reports

Enhancing Cyber Resilience through Insurance: Revisiting Anti-Bundling Regulation

Sophia Mauro and Taylor Grossman

viewpdf

Op-ed

ROOST Reminds Us Why Open Source Tools Matter

view

Reports

Navigating AI Compliance, Part 2: Risk Mitigation Strategies for Safeguarding Against Future Failures

Mariami Tkeshelashvili, Tiffany Saade

viewpdf

Reports

Deterring the Abuse of U.S. IaaS Products: Recommendations for a Consortium Approach

Steve Kelly, Tiffany Saade

viewpdf

Contribute to our Library!

We also welcome additional suggestions from readers, and will consider adding further resources as so much of our work has come through crowd-sourced collaboration already. If, for any chance you are an author whose work is listed here and you do not wish it to be listed in our repository, please, let us know.

SUBMIT CONTENT

AI and Human Decision Making: AI and the Battlefield

Institute for Security and Technology, Center for Global Security Research

SUMMARY

As the 21st-century geopolitical balance shifts in uncertain ways, there is an increasing eagerness to deploy AI technologies into both the physical and digital battlefields to gain both tactical and strategic advantage over adversaries. However, the nature of increasingly powerful and unpredictable AI demands a measured and balanced approach to deploying these tools before the limitations, risks, and vulnerabilities are fully understood and addressed. Indeed, these technologies may not currently be “ready for primetime”, on a number of levels. We begin this discussion – meant to be a series of posts on this domain of issues – focused on the following. This initial paper is based off of numerous small-group workshops and ongoing engagement with the AI research community in the San Francisco Bay Area. This initial paper is based off numerous small-group workshops and ongoing engagement with the AI research community in the San Francisco Bay Area: Current AI capabilities remain limited to narrow, well-defined domains. ​ The “black box” nature of state-of-the-art AI/ML algorithms gives limited insight as to their decision-making processes – and conclusions Deploying AI’s could preempt ethical considerations that have yet to be fully understood, identified, or agreed upon, and is in the potential context of an industry-driven race to the bottom

download pdf