For two decades, SHIFT has guided clients through communications challenges as automation and artificial intelligence entered, disrupted and transformed their markets, from healthcare to manufacturing to transportation and beyond. Now, it has finally come for us: AI in public relations is here.
You’ll have read the headlines. Generative AI (genAI) burst onto the scene in late 2022 with the public availability of OpenAI’s ChatGPT, one of many apps that create text, images, videos and audio content based on its learnings from massive data stores. Adoption was swift: 100 million users in two months. Compare this to the 4.5 years it took Facebook to hit that milestone. Even TikTok needed nine months.
At SHIFT, we’ve often been an early adopter of new technologies in communications (from social media to Google Analytics). With genAI, we were quick to recognize the inflection point our industry faces. As responsible communicators, it is our job to both help clients navigate how it impacts our discipline and to use it to make ourselves more effective. We take the word “responsible” seriously, and we have spent hours deliberating if, when and how our staff should use genAI tools.
With that in mind, SHIFT has added genAI to our toolbox and encouraged its use, though with guardrails. They are as follows and are based on those internal discussions, as well as on outside perspectives, including the PR Council Guidelines on Generative AI.
Our guardrails for generative AI in public relations
Keep private information private.
We are privileged to have the trust of clients who rely on us to give counsel on news and activities that have the potential to move markets. genAI apps are ultimately owned by private companies that may use information entered into their queries to train future versions. As such, we are mindful of what we upload. Entering proprietary information into a database not under client or agency control breaks that trust — and could have legal ramifications.
- How we use it: for general research on topic areas.
- How we don’t: to write press releases, talking points or other materials that are not already in the public domain.
Use it transparently.
Just as you would cite the source of a data point or credit a media outlet for breaking a piece of news, communications professionals need to be transparent when genAI has influenced their content.
- How we use it: to inform and ideate.
- How we don’t: position AI-derived copy as our own. We under no circumstance present the output of a genAI query to a client as final content.
Understand its bias.
Bias in AI has been an issue since its earliest days, and genAI tools are only as good as the databases on which they are trained. The PR Council’s guidelines have a dozen questions to help suss out any potential bias while using generative AI in public relations.
- How we use it: with an eye toward our firm’s ethical standards. SHIFTers at every level adhere to them.
- How we don’t: by ignoring the fact that these models are based on as much “bad” information as they are “good” data.
Fact-check everything.
genAI models are trained on massive databases that encompass books, articles and websites. This means they can include misleading or false information — including presenting as fact information they have made up wholesale.
- How we use it: to jumpstart research. It can “ELI5” anything from quantum mechanics to the cause of political polarization.
- How we don’t: accept everything it generates as the absolute truth or use it as an only resource. We double-check stats and sources via search or original materials. (AI hallucinations are real. For a fun exercise, ask ChatGPT to write your bio. Apparently, I graduated from Boston University and have been published in TechCrunch; neither is true.)
Accelerate work but know its limits.
One of our leaders refers to ChatGPT as an “intelligent idiot.” While it can produce passable output, it largely lacks the context, creativity and authenticity of human-produced content. The latter is required for a pitch to stand out in a reporter’s crowded inbox, to compel an audience to read beyond the first two lines of an article, or to create thumb-stopping content.
- How we use it: to spark creativity and speed up the cycle between ideation and execution. genAI can help marketers and communicators brainstorm new ideas and perspectives. It’s proved to be an excellent thesaurus and helped us break writer’s block. Our Creative Director has used DALL·E to turn early ideas into initial design sketches (I can just sense an upcoming opinion piece from him on using genAI as a launch point for original work).
- How we don’t: as a substitute for original thinking. Our value to clients lies in our experience and expertise; it always will.
genAI is just the latest arrows in our communications quiver. Just as we have seen in other industries, AI in public relations will free communicators from some of the duller tasks or grease the wheels in other processes, allowing us to focus on more creative and strategic parts of our discipline — the parts that ensure what we do delivers real performance and impact.
What’s a Rich Text element?
The rich text element allows you to create and format headings, paragraphs, blockquotes, images, and video all in one place instead of having to add and format them individually. Just double-click and easily create content.
The rich text element allows you to create and format headings, paragraphs, blockquotes, images, and video all in one place instead of having to add and format them individually. Just double-click and easily create content.
Static and dynamic content editing
A rich text element can be used with static or dynamic content. For static content, just drop it into any page and begin editing. For dynamic content, add a rich text field to any collection and then connect a rich text element to that field in the settings panel. Voila!
How to customize formatting for each rich text
Headings, paragraphs, blockquotes, figures, images, and figure captions can all be styled after a class is added to the rich text element using the "When inside of" nested selector system.
