Research Library

Improving the Explainability of Artificial Intelligence: The Promises and Limitations of Counterfactual Explanations

By: Alexander Asemota, Improving the Explainability of Artificial Intelligence: The Promises and Limitations of Counterfactual Explanations

A new white paper from the Center for Long-Term Cybersecurity explores diverse approach to explainable artificial intelligence (xAI), focusing on counterfactual explanations, or CTEs. The paper, “Improving the Explainability of Artificial Intelligence: The Promises and Limitations of Counterfactual Explanations,” was authored by Alexander Asemota, a 2023-2024 AI Policy Hub Fellow.

Benchmark Early and Red Team Often: A Framework for Assessing and Managing Dual-Use Hazards of AI Foundation Models

By: Anthony Barrett, Krystal Jackson, Jessica Newman, Nada Madkour, Evan R. Murphy, Benchmark Early and Red Team Often: A Framework for Assessing and Managing Dual-Use Hazards of AI Foundation Models

Authored by Anthony M. Barrett, Krystal Jackson, Evan R. Murphy, Nada Madkour, and Jessica Newman, this report assesses two methods for evaluating the “dual-use” hazards of AI foundation models, which include large language models (LLMs) such as GPT-4, Gemini, Claude 3, Llama 3, and other general purpose AI systems.

The Transaction Costs of Municipal Cyber Risk Management

By: Rowland Herbert-Faulkner

This white paper, The Transaction Costs of Municipal Cyber Risk Management, brings to light the transaction costs associated with municipal cyber risk management, including the costs of searching for information, coordination between parties, drawing up and enforcing contracts, negotiation, inventory and monitoring, and compliance and enforcement. The paper was authored by by Rowland Herbert-Faulkner, a PhD Candidate in the Department of City and Regional Planning at UC Berkeley whose dissertation research focuses on technology governance at the municipal and regional scales.

CLTC 2023 Annual Report

Click below for an overview of highlights from the UC Berkeley Center for Long-Term Cybersecurity’s activities in 2023. Download a PDF version of the 2023 Annual Report:

BreachProphet

By: Ravi Patnala, Mayank Saxena, Meenakshi Sriraman

BreachProphet aims to address critical gaps in current cybersecurity solutions, particularly the lack of an effective mechanism to map and predict risks stemming from vulnerabilities for businesses by…

LLM-Powered Spear Phishing Detection Solution

By: Caleb Albornoz, Jeff Chang, Gus Maldonado, Charlie Zhang

Problem Statement “Imagine a bustling city street, alive with people going about their daily life. In this action-packed environment, a skilled pickpocket moves with stealth, blending anonymously and…

Representing Privacy Legislation as Business Risks

By: Andrew Chong, Richmond Wong

For this CLTC white paper, researchers Richmond Wong and Andrew Chong used Form 10-K documents — annual regulatory reports for investors that publicly traded companies must file with the U.S. Securities and Exchange Commission (SEC) — to analyze how nine major technology companies assess and integrate the business risks of privacy regulation like the EU’s General Data Protection Regulation (GDPR), the California Consumer Privacy Act (CCPA), and the California Privacy Rights Act (CPRA).

Cybersecurity Futures 2030: New Foundations

To better understand how diverse forces are shaping the future of cybersecurity for governments and organizations, the Center for Long-Term Cybersecurity (CLTC), the World Economic Forum Centre for Cybersecurity, and CNA’s Institute for Public Research collaborated on “Cybersecurity Futures 2030: New Foundations,” a foresight-focused research initiative that aims to inform cybersecurity strategic plans around the globe.

A Taxonomy of Trustworthiness for Artificial Intelligence

By: Jessica Newman

The National Institute of Standards and Technology (NIST) has developed an AI Risk Management Framework (RMF) intended to promote trustworthy artificial intelligence (AI). In this report, we introduce…

From Policy To Pixels: Strategic UX Design and User Support for GDPR Implementation

This piece originally appeared on the Superbloom Blog here: https://superbloom.design/learning/blog/from-policy-to-pixels-strategic-ux-design-and-user-support-for-gdpr-implementation/ By: Susan Kennedy, Ame Elliott, and Georgia Bullen Introducing our new report “From Policy to Pixels: Strategic UX…