Data privacy and overcoming employee resistance are the two main sources of anxiety for companies investing in artificial intelligence.

According to our 2024 report, Generative AI in the B2B Marketplace, 81% of B2B businesses are already investing in AI, with 79% anticipating increasing their AI budgets within the next year. Yet these same companies have the same anxieties about investing. 

AI, at its heart, is a tool to help solve problems — like a hammer, or pair of scissors. But just as we teach kids not to run with scissors, we have to set up guardrails around AI to make it both safe and useful.   

Data privacy and employee resistance are two running-with-scissors risks. Without pragmatic, thoughtful practices, companies risk harming buyer-seller relationships, employee morale and overall productivity.  

What can supply chain leaders do to mitigate these risks? Make an AI alliance around key customer use cases.

Customers Like AI ... Sometimes  

The B2B buyer report revealed that 65% of buyers found AI-powered chatbots to be moderately or extremely helpful when it came to purchases. However, only 48% said that the chatbots provided all the information they needed — meaning that 52% of buyers still needed to speak to a customer support agent to find their answers.

Additionally, it showed that 54% of buyers prefer live chat support when purchasing on a web store, the most preferred form of support. According to Forbes, 86% of buyers prefer interacting with humans over AI.  

AI chatbots are beneficial, but they have room for improvement, so supply chain executives should understand the pros and cons and deploy them with caution. One solution is to provide chatbots, but then transfer buyers to agents when the chatbot is no longer able to solve their challenge. This would allow for cost savings and speed increases of AI, while still providing the live chat ability that buyers still need in many cases.  

With the rise of ChatGPT, the popular conception of AI has been that of a program that a user prompts, then receives a response. But there are other forms of AI that benefit customers.  

A key example is guided selling — a dynamic form of product search that asks users a series of questions to help winnow down the catalog to a manageable set of products that meet their exact specifications. The tool helps buyers find products faster, and reduces order errors. Additionally, it can free up sales and customer support teams from tedious customer inquiries on relatively simple tasks, allowing them to focus on more proactive and large-scale projects.  

This tracks with findings that 69% of companies believe AI will make their search and discovery processes better, and 86% intend to use more AI for that purpose in the next year.

Data Privacy and Security Worries 

A key risk that companies have identified is a lack of data privacy or security. Because AI models are trained on an enormous amount of data, there is a risk that sensitive data could be leaked or otherwise extracted. As an example, a savvy prompter could prompt a chatbot, trained on customer data, to divulge confidential data such as payment information. Since GenAI is still relatively new, privacy and security guardrails are still being developed. It’s not quite the Wild West, but it’s not yet as safe and regulated as air travel either. 

Companies that fail to develop guardrails around ethical and effective AI use risk both reputational damage and legal danger.  

As an example, an AI-based recruiting software explicitly rejected a candidate due to age,  violating anti-discrimination statutes and opening the company up to liability. Similarly, New York City’s MyCity chatbot gave advice that was explicitly illegal, saying that restaurants could serve rodent-bitten food or that employers could take a cut of employee tips.  

Those are expensive mistakes for any supply chain executive to make.  

AI has the opportunity to free workers in the supply chain industry from the drudgery of repetitive tasks, while also helping customers solve their problems faster than ever. However, it risks giving false information, divulging sensitive data and causing customer heartburn that ends up putting more stress on your workforce.  

That’s why it’s critical to view AI as a tool to assist your business and customers. It’s best used to augment and supplement, with the understanding that buyers both appreciate and prefer human support. 

Don’t run with scissors. Cut with precision.

Arno Ham is B2B e-commerce evangelist at Sana Commerce.