AI disclaimers should be a PR industry standard

AI is in the news every day.

The week I wrote this, the German artist Boris Eldagsen turned down the prize for the creative open category at the Sony world photography awards because his winning entry was generated entirely from AI.

The next day, a song featuring AI-generated vocals purporting to be Drake and The Weeknd went viral before being removed from streaming services. There have also been stories about criminals using voice cloning software to scam family members and businesspeople out of money. A German magazine even published a fake AI-generated interview with F1 legend Michael Schumacher.

The release of ChatGPT has brought to the surface the negative, damaging, and downright nefarious impacts of AI. An open letter by over 1000 experts and tech leaders warned about the potential risks to society and humanity of human-competitive AI. Some have even suggested in could spell the end of all humanity.

Embracing AI

Doomsday predictions aside, I believe AI should be embraced in the PR industry. It can solve problems we didn’t even know existed.

This morning I came across an ad on LinkedIn for a coverage report tool that claimed to save PR practitioners from ‘coffee-fuelled all-nighters’ spent creating coverage reports. While I can only dream of securing so much coverage it takes all night to paste it into a report, the underlying message of the ad resonated with me. How can we take the mundane, repetitive, and downright boring out of PR?

AI is a big part of the answer. AI tools can probably carry out most of the functions the coverage report tool mentioned above can. Possibly even more. CIPR recently found that there are roughly 5,800 technology tools with potential applications in the public relations industry – 5,800 opportunities to make our lives easier.

I wanted someone to validate my argument, so I turned to ChatGPT. I asked it why AI should be used in PR. It listed six reasons – automation of repetitive tasks; data analysis and insights; improved targeting of audiences; personalisation of communications; predictive analysis of events; and crisis management.

All accurate answers. But also very generic. Which is exactly why AI should only ever play a supporting role – because it still requires human finesse. Sending anything entirely generated by AI to a client without sense checking it is a fool’s errand.

I recently heard about a political journalist who relied entirely on ChatGPT to prepare for an interview with a minister. It saved the journalist some valuable time and provided him with extensive background information about the politician’s policies and achievements.

The only problem was that most of the information was wholly inaccurate, which led to a very embarrassing interview. Asking a minister why they introduced a policy that they’d never heard about is like sending a client a piece of coverage that doesn’t mention them.

So we don’t need to worry about AI stealing our jobs. For now.

Although it’s not just media monitoring and reporting that AI can help with. It can also play a supporting role for strategy and positioning, press releases, photography, video production, podcast production, social media graphics and blog posts.

Attracting talent

But perhaps the biggest way it can help our industry is attracting the best possible talent.

I know from experience it’s hard to find good candidates with solid experience in PR. The competition for talent amongst agencies is fierce. We need to do more to more to market PR as a creative and disruptive industry if we are to attract talent from other industries. 

If we can use AI to delegate mundane tasks like media monitoring and audio transcribing, it allows us to invest more in creative skills and offer more in-house services such as studio production and graphic design. It enables us to become a full-spectrum creative industry – not one based on press office activity and obsessed with coverage reports and readership numbers.

But what are the risks of using AI in PR? I asked ChatGPT why it shouldn’t be used in the industry.

It gave me six reasons – lack of human connection; lack of context and judgment; potential bias and discrimination; overreliance on technology; and privacy and security concerns.

Again, these are accurate. But these potential issues pale in comparison with the reputational risk to the industry if we don’t tell clients when and how we use it.

Let’s be honest, PR already has a somewhat dubious reputation to those unfamiliar with the industry. Many people think PR agencies are only called into action to cover up a corporate crisis or sweep some unsavoury truths under the carpet.

And if agencies start charging standard rates for services carried out using AI capabilities, we could face our own self-inflicted PR disaster.

Disclosing to clients how we use AI tools to support with communications activity isn’t an honourable act – it should be the baseline expectation and industry standard. I think all agencies should publish an AI disclaimer on their websites and direct clients to this when they start work with them.

Sam Patchett is client director at communications agency Profile, which has produced its own AI disclaimer.