Sebastiaan Verhaar

“Open the pod bay doors, HAL!” I’ll never forget the chill that ran over me as I saw Dave’s anger turn to dread in “2001: A Space Odyssey”. Dave, the movie’s shrewd and unrelenting astronaut, was being hunted down by the AI-powered computer character HAL, who was supposed to assist and protect him.

There has been no shortage of killer robots throughout pop culture history: HAL in 2001 a Space Odyssey, Skynet in the Terminator, PAL in Mitchells vs. The Machines (check that one out if you haven’t!) — it’s almost as if we, as humans, are presupposed to the belief that robots will turn against us.

Enter generative AI (GenAI), which captured public imagination at all levels, from business to academia to pop culture, starting in the fall of 2022. Suddenly, here was a piece of technology that could create — on demand — stories, emails, presentations, songs! We were promised that the humdrum of churning out generic content was a thing of the past!

The truth, of course, is less dramatic than what movies would suggest and less glamorous than what we were promised.

Still, GenAI has upended our world. For B2B businesses, the use cases range from streamlining customer service to smart search to data analysis to automated workflows to enhanced decision-making.

While our businesses, so far, have yet to face the nightmare of a rogue AI telling us, “I can’t do that for you, Dave,” when we ask it to “open the spreadsheet, HAL;” we must be honest that there are ethical quandaries and risks we must navigate.

Remember: AI is a tool. Like any other tool, be it a shovel, a hydraulic press, or a Xerox machine, it is up to us to use it ethically.

Here’s how to navigate the ethics of GenAI in a B2B space.

GenAI: Everyone else is doing it.

WBR insights surveyed leading B2B businesses on their use of GenAI in the workplace. A supermajority of 81% reported that they were already using AI, and 79% expect to increase their AI investments this year. Their most important use cases were in automating tasks (77%), analyzing data (75%), enhancing decision making (66%), and generating internal communications (54%).

These tools are powerful — remember the first time I received a GenAI summary of a video call I’d been on. Instead of a word-for-word transcript, I received a breakdown of key points with timestamps that would take me to the relevant parts of the call. GenAI tools like these immediately prove their efficiency; it’s not hard to see why companies adopt them readily.

However, the survey also revealed some blinking yellow lights: 74% of respondents are concerned about AI errors, while 68% have concerns about ethics and data privacy.

B2B leaders need to take these concerns seriously.

The peril of AI hallucinations

One of the largest ethical concerns identified is the risk that GenAI could introduce false information into its generated content. For example, GenAI could be trained on your past sales data, generate inaccurate synthetic data, then train on this synthetic data while magnifying errors. This process is called “AI hallucinations,” where an AI-generated error becomes more prominent and widely disseminated.

For B2B businesses, this could prove challenging when it comes to customer-facing chatbots, generated reports based on predictive modeling, or even improperly generated AI product descriptions.

The solution relies on employees fact-checking and reviewing all AI-generated content before it is publicly or internally released.

The risk of bias 

Another significant ethical concern is the risk of bias in GenAI. In 2023, Bloomberg ran a series of tests using image generator, Stable Diffusion. They prompted the AI to create images of people working in various fields — architects, doctors, house keepers, fast food workers.

The results? White faces were overrepresented in the high-paying jobs (like architects) while Black and Brown faces were overrepresented in lower-paying jobs (like fast food workers). The bias arises from the data on which Stable Diffusion was trained; it drew from images that were biased to showing certain ethnicities in certain positions, and it then magnified this bias in its output.

Bias can be tricky, because it is often hard for us to identify our own blind spots. For B2B businesses, it is best to review each piece of generated content with a keen eye toward identifying bias.

Is that image copyrighted? 

Copyright in the GenAI space is an ethical and legal quagmire. Artists and authors have initiated class-action lawsuits against AI companies, accusing them of scraping images and texts for training AI models —without compensation.

B2B businesses can take steps to mitigate legal risks by choosing AI tools that have been trained on royalty-free or public domain data.

GenAI in B2B: set appropriate parameters for HAL

In 2001: A Space Odyssey, HAL malfunctioned because he received two conflicting sets of commands: never lie to the crew; keep their true mission a secret from them. The result: he malfunctioned.

For us, the immediate concerns of a GenAI malfunction aren’t as severe as killer AI, but they are still significant. Inaccurate and biased content could lead to significant business setbacks.

The solution, therefore, is to “box HAL in.” Develop and put consistent processes in place for AI use within your workplace. Identify key tasks that you want AI to accomplish, have employees review the content before it is released, and choose AI programs that are trained on business-friendly data.

At the speed GenAI is proliferating, it can feel like you need to jump on the newest AI tool immediately for fear of missing out on a competitive advantage. Temper that. Proceed purposefully, judiciously, and (most importantly), ethically.

About the author:

Sebastiaan Verhaar is the CEO of Sana Commerce, an ecommerce platform engineered for B2B.

Sign up

Sign up for a complimentary subscription to Digital Commerce 360 B2B News, published 4x/week. It covers technology and business trends in the growing B2B ecommerce industry. Contact Mark Brohan, senior vice president of B2B and Market Research, at [email protected]. Follow him on Twitter @markbrohan. Follow us on LinkedIn and be the first to know when we publish Digital Commerce 360 B2B News content.