What’s your agency framework for generative AI? Here’s how we created ours

Just look at Samsung. The brand confirmed in May that it had placed an employee-wide ban on ChatGPT after the upload of sensitive data to the platform. This is an inevitable outcome when companies haven’t yet figured out their policy on generative AI. But with such transformational new technology, how do you even go about creating that policy in the first place?

If, like me, you’ve spent a fair bit of time using ChatGPT and Bard, you’ll have encountered the hallmarks of current-gen artificial intelligence text generation. This includes bland language, repetitive phrases, and sometimes dubious factual statements that you’d be mad to down on paper. But likewise, you may have discovered path unblocking suggestions to approach a communications challenge, or (with enough prompting) a solid outline to get stuck into your next client op-ed.

During Q1 our agency team was also exploring generative AI and making similar observations, focusing on how to automate the more routine tasks of PR agency client servicing and free up time to, for example, give deeper strategic counsel. At this time, we also saw tech media outlets publishing their editorial stances on generative AI. This was both bold and inspiring, especially considering the speed of online debate around AI and the future of journalism and content creation more broadly.

So as an agency, we concluded that if our media friends are boldly and transparently publishing their ground rules relating to generative AI, won’t our clients soon expect us to do the same?

This was why we recently published our framework for how we use generative AI at Bud.

It took us over two months to publish. Step one was to get the entire company together for a workshop, with everyone presenting ideas on acceptable use cases, as well as practical and ethical challenges. Step two involved a workable draft, which was then mulled over and tweaked by our entire team. Step three was to figure out the best client-facing comms approach.

The framework is a work in progress. We’re very open that this will need to be built upon (and edited) over time to reflect how current-generation generative AI evolves and how we intend to leverage it. For now, our framework is an exhaustive representation.

It’s worth remembering that we don’t know what we don’t know. We think this is an effective set of use cases, but we’re very open to re-think our approach if and when we learn otherwise.

The biggest questions we asked ourselves before putting this framework into use were:

• Does this respect the trust our clients have placed in us?

• Does it represent our clients’ best interests?

• Will it positively impact the overall quality of our work?

• Will it give us greater bandwidth to flex our attributes as top PR pros?

As an agency leader, my hope is that generative AI will allow us to deepen our relationships with clients, freeing up more time to address the challenges and opportunities at the heart of their business. Thanks to this incredible technology, we’re about to do the best work of our careers, providing we create and adhere to a clear set of ground rules. Meanwhile, for those not investigating generative AI at all, the biggest existential business threat could soon be the people leveraging this technology (not the technology itself).