generative ai governance

Don’t be scared of generative AI – bet on AI governance instead

With the great promise of generative AI technologies come great risks. But completely banning them from being used in your company is a bad idea – you should focus on establishing clear rules and procedures instead. Embrace AI governance.

Back in November 2022, the up-and-coming engineers at OpenAI released a demo that, they thought, wouldn’t make much of a splash. Instead, their brainchild turned out to be one of the most impactful digital tools ever. We’re talking about ChatGPT of course, and it’s safe to say that almost anyone involved in business has heard of it in some capacity. 

ChatGPT was AI’s breakthrough moment, and turned generative AI, specifically, into the darling of the tech industry. Generative AI allows you to effortlessly create content (whether text, images, videos, or sound) and pingpong ideas back-and-forth. ChatGPT’s main contribution was really that it made interaction easy and natural, allowing even people with lower digital literacy to use it without a hitch. 

The business world has also recognized the potential of GenAI, as executives and employees alike have been using it to draft emails, create roadmaps and plans, write reports, generate marketing material, and much more. 

However, with its great opportunities come great risks. For example, since GenAI ‘learns’ from massive datasets scraped from the internet, it also contains a lot of copyrighted material. Because GenAI tends to run its learning dataset through a blender and then reassemble it into new material, it’s not easy (if not impossible) to trace back materials to their source. Also, there are significant privacy and data security issues with these tools. They can contain sensitive personal information that can potentially be accessed either indirectly or even, with some creative ‘prompt engineering’, directly. 

A total ban is bad risk management

Some organizations, like the Dutch public administration, have decided to impose a total ban on GenAI tools like ChatGPT to steer clear from these IP and privacy risks. However, it would be a much more effective strategy to focus on AI governance: on creating the right rules, processes, and structures in your organization to manage these risks while embracing the gains that GenAI brings. 

Lay down clear rules regarding the data you enter and how you use produced material. For example, you can tell employees to only generate internal emails (not external ones) and that they cannot use sensitive data. If you add quality assurance checks, you’re extra safe. Moreover, with the help of AI experts like us at Sparkle, you can easily create a secure environment where you can work with your own data without any issues. Or tell your employees to never use GenAI outputs in a public context such as social media, and that they should double-check all factual claims. AI governance essentially means you’re safe from the risks, but still getting all the gains from GenAI. 

Trust your people, embrace AI governance

A knee-jerk reaction like a total ban harms not just your productivity, but also your organization’s image and culture. As more and more people know about responsible AI governance, business partners and customers might start to wonder whether you don’t have the organizational acumen to create the rules and policies you need to use it safely. That doesn’t reflect well on your organization’s leadership. 

Furthermore, banning GenAI implies you don’t trust the digital skills of your employees. If you train your people well, they’ll know that ChatGPT doesn’t give you trustworthy factual information or that they shouldn’t blindly follow its advice. Your organizational culture should reflect the trust you have in the capabilities of your employees, and a total ban throws all of that out the window. Creating training programs and awareness sessions on (generative) AI is key. 

In sum: you should embrace generative AI, but only do that with the proper guidance and AI governance. At Sparkle, our expert AI consultants can help you capture the full value of GenAI without exposing you to any of the risks. We can help you create responsible processes, set up secure environments, safely leverage your data, and set up training and awareness programs.

Do you want to jump on the AI train without taking on risks? Get in touch with our team via Koen Mathijs.

ABOUT THE AUTHOR

Jens Meijen

Jens Meijen

PhD at KU Leuven & Consultant at Sparkle

Scroll to Top