Generative AI and Procurement: a match made in heaven, or a cause for concern?

As we consider the future of procurement, there's no denying that topics like digitalisation, supplier collaboration, and sustainability are often at the forefront of discussions. However, in recent months, the use of ChatGPT and generative AI has become a hot topic, with many wondering about its impact on procurement.​

The question we must ask ourselves is, how can procurement professionals leverage Natural Language Processing (NLP) AI to maximise performance and value while also being ethical and considerate? The benefits are clear. Automating and streamlining processes through AI can free up capacity and allow procurement professionals to focus on more strategic tasks, such as quickly analysing data to identify trends, patterns, and opportunities, resulting in more informed decision-making.​

Where we see the greatest potential for AI in procurement is in managing risk proactively. With uncertainty on the rise due to geopolitical pressures and natural disasters, having access to accurate and detailed information at the touch of a button will be crucial in mitigating and managing risk, and can help level the playing field for small and medium-sized enterprises. Examples include leveraging​:

1. Textual data analysis: allowing organisations to analyse large amounts of unstructured textual data, such as news articles, social media posts, and customer feedback.​
2. Contract analysis: supporting organisations to analyse contracts and identify potential risks in the terms and conditions. For example, clauses that may increase their exposure to liability or regulatory risks.​
3. Fraud detection: by analysing large amounts of data, such as email and chat messages, helping to identify patterns in communication that may indicate fraudulent behaviour and flag areas for further investigation.​
4. Compliance: assisting organisations to be compliant with regulations and standards by analysing policy documents and reports and identifying areas of non-compliance in order to take corrective action.​
5. Predictive modeling: used to build predictive models that identify potential risks and provide early warnings. For example, analysing news articles and social media posts to predict potential market risks and provide early warnings.​

However, one major challenge that must be addressed in the use of Natural Language Processing AI is the risk of the spread of misinformation due to biases. This is because NLP AI systems are trained on large amounts of data and if that data is biased, it can result in biased and incorrect predictions and decisions causing reputational issues for organisations. An example of this was highlighted recently with Google's AI bot, Bard, providing an incorrect answer during an advert on Twitter, resulting in its parent company, Alphabet, losing £82 billion in market value and raising concerns about its capabilities.​

This example highlights perfectly the damage AI can have when used improperly or without appropriate consideration. Hence, it's imperative that when applying this technology, we exercise caution and take a prudent approach, as despite its advantages, there are still persistent risks that need to be considered. Despite this, as Natural Language Processing AI continues to improve and procurement professionals better understand how to apply it, there is no doubt that it has the potential to be a true accelerator for delivering greater value.

Author

Dan Bone

Senior Consultant

dan.bone@bearingpoint.com