Important Dates
- Submissions due: 1 May 2024
- Publication: November 2024
In the past decade, we have seen exponential growth in intelligent and autonomous systems development and deployment. Along with this fast proliferation, we are witness to continued rise in reports of autonomous learning system failures, malfunctions, and undesirable outcomes. Multiple efforts to log these failures have also been initiated.
We learn more from analyzing failures in engineering than by studying successes. There is significant value in documenting and tracking AI failures in sufficient detail to understand their root causes, and to put processes and practices in place toward preventing similar problems in the future. Efforts to track and record vulnerabilities in traditional software led to the establishment of National Vulnerability Database, which has contributed toward understanding vulnerability trends, their root causes, and how to prevent them.
Computer magazine is soliciting papers for a special issue on AI Failures: Causes, Implications, and Prevention. This special issue will explore AI failures, from early systems to recent ones. Papers should discuss the causes of the failures, their implications for the field of AI, and what can be learned from them.
Topics of interest include, but are not limited to:
- Specific AI systems that have failed
- Autonomous vehicles
- Diagnostic Systems
- Medical devices
- Decision aiding tools
- Recommendation systems
- Robotics
- The causes of AI failures
- Inadequate training data
- Testing failures
- Human interaction with AI/machines
- Adversarial attacks on AI systems
- Transfer learning problems and evolution of use, environment
- The implications of AI failures
- Trust/acceptance
- Societal and legal implications
- Quantification of loss from AI failures
- Economic impact
- Research directions
- Regulatory Issues
- What can be learned from the failures
- Root cause analysis
- Fault tolerance techniques
- Testing methods and adequacy
- Importance of assurance metrics and methods
- How to avoid AI failures in the future
- Documentation and reporting of failures
- Safety/security analysis methods for AI/ML
- Explainability integration
Submissions should be original and unpublished.
Submission Guidelines
For author information and guidelines on submission criteria, visit the Author’s Information page. Please submit papers through the ScholarOne system and be sure to select the special issue or special section name. Manuscripts should not be published or currently submitted for publication elsewhere. Please submit only full papers intended for review, not abstracts, to the ScholarOne portal. If requested, abstracts should be sent by email to the guest editors directly.
Questions?
Contact the guest editors at co11-24@computer.org.
