This chapter explores the potential of leveraging generative artificial intelligence (AI), specifically ChatGPT, for assessing corporate sustainability. We examine ChatGPT’s reliability in sustainability scoring and investigate the factors contributing to discrepancies between AI-generated scores and a composite benchmark from established traditional data providers, namely Refinitiv, Bloomberg, and S&P500. Our methodology includes automating ChatGPT to evaluate companies based on emissions, innovation, and resource use, comparing these scores to a composite of the three benchmark scores (RBS), and assessing the associated misclassification risk. We also explore whether ChatGPT’s suitability varies based on firm-specific characteristics such as size, geographic location, industry sector, age, and profitability. Our findings reveal that ChatGPT can generate sustainability assessments, but with limitations. The AI shows a high non-response rate and significant differences in score distributions compared to RBS. ChatGPT tends to produce higher, more concentrated scores and struggles with firms exhibiting lower environmental performance. Despite these challenges, an association exists between ChatGPT and RBS, indicating partial overlap in informational content. Certain firm characteristics, including larger size, European location, and older age, enhance the reliability of ChatGPT’s assessments. These results emphasize the need for more openly available and regulated sustainability data, as well as the importance of continued innovation and improvement in AI technology to enhance the coverage and accuracy of such assessments. This research contributes to a new approach to assessing corporate sustainability, where advanced technologies and open, unbiased data support businesses in their sustainability journey.