When to Trust AI: Advances and Challenges for Certification of Neural Networks

TL;DR

This paper reviews recent advances and challenges in certifying neural networks to ensure their safety, reliability, and trustworthiness in critical applications like autonomous systems and healthcare.

Contribution

It provides a comprehensive overview of current certification techniques for neural networks and discusses future challenges in ensuring AI safety and explainability.

Findings

Survey of existing certification methods

Identification of key challenges for neural network safety

Discussion of future research directions

Abstract

Artificial intelligence (AI) has been advancing at a fast pace and it is now poised for deployment in a wide range of applications, such as autonomous systems, medical diagnosis and natural language processing. Early adoption of AI technology for real-world applications has not been without problems, particularly for neural networks, which may be unstable and susceptible to adversarial examples. In the longer term, appropriate safety assurance techniques need to be developed to reduce potential harm due to avoidable system failures and ensure trustworthiness. Focusing on certification and explainability, this paper provides an overview of techniques that have been developed to ensure safety of AI decisions and discusses future challenges.