Here in Scotland, 30 November is celebrated as St Andrew’s Day and is now an official public holiday. However the date will be marked by many for other reasons. It will mark exactly one year since ChatGPT (Chat Generative Pre-trained Transformer) launched as the first artificial intelligence (AI) tool accessible to the public around the world. Since then, AI has captured the imagination of more than 100 million users globally and is starting to impact the business world as well.
What does it mean for PR and public affairs? PA Cooperative recently attended a great training session organised by the Chartered Institute of Public Relations in Edinburgh and this article builds on the lessons learned there.
Introduction
From image creation to copywriting, AI is slowly but surely making its way into our professional lives. But when it comes to public affairs and media relations, what are the ethical implications of using AI? With confidentiality, data protection, and sensitive communications to consider, is it acceptable to harness AI within your public relations? Let’s take a look at the issue a little more closely.
Harnessing AI algorithms for PR and media relations
AI algorithms identify patterns and themes in content. These powerful tools can analyse vast amounts of data, far beyond human capabilities, and generate outcomes in seconds.
Algorithms can be incredibly useful in supporting PR professionals. They can detect deep fake videos and identify emerging issues by analysing news and social media trends, enabling early detection of issues and timely responses.
AI can also search a whole social media platform or a variety of news outlets to assess live public opinion of a person or organisation.
Machine-learning AI algorithms can create content too, by transcribing audio and video files to text in minutes – on platforms such as Otter.ai. These platforms enable much faster content production than a human working alone, although AI transcriptions are never completely accurate, with varying results depending on the audio quality and speakers’ accents. One video that PA Cooperative worked on threw up some very “interesting” AI-generated captions from an interviewee with a heavy Irish accent!
This all sounds beneficial — so what are the ethical concerns around the use of AI algorithms?
Concerns exist regarding the management of personal information used in analysis. The Information Commissioner’s Office (ICO), which provides guidance on artificial intelligence and data protection laws, recommends that organisations clearly communicate how they handle data in alignment with GDPR. Specifically, they advise transparency about why personal data is gathered, the length of time before deletion, and the parties with access privileges.
Adhering to proper data handling principles is important, especially given public wariness over how their information feeds into AI systems. By following ICO suggestions around purpose specification, retention periods, and access rules, organisations can ethically and legally leverage personal data to train algorithmic models.
So, if your data handling practices are up to scratch, it’s generally ethical to use AI algorithms to support human PR professionals’ work.
Using generative AI (GenAI) within public relations
Generative AI tools, like ChatGPT or Claude, produce entirely new pieces of content. GenAI tools can research and draft content pieces in seconds, saving time and ensuring more thorough research than would otherwise be possible. However, using generative AI in public relations is where things begin to get murky.
Most GenAI tools are not advanced enough to write about delicate or controversial topics in a sensitive way, and using GenAI for PR is delicate from an ethical standpoint. Organisations risk being seen as treating issues flippantly and without due care by using AI to create their responses, especially if the content is perceived as insensitive. AI models have also been proven to exhibit bias, since they take on human biases from the material used to train them.
More importantly, AI content isn’t always accurate, and the issue of ‘hallucination’ is a major one for the AI community. Hallucination occurs when a large language model (LLM) – like a generative AI chatbot – perceives patterns or objects that are non-existent or imperceptible to humans, creating outputs that are inaccurate.
Any PR content generated by AI must therefore be thoroughly checked and edited by a human before being published. The organisation should also be prepared to be transparent about using AI, to avoid reputational damage.
Ethical data handling and confidentiality issues are also relevant here. Some organisations use AI to respond to emails and messages, and for PR firms, it can be a tempting idea to get through volumes of media enquiries.
However, anything submitted to GenAI is processed and becomes part of the AI’s learning, which the person must be aware of and able to consent to, such as when interacting with a chatbot.
Therefore, many schools of thought believe it is unethical to use GenAI to respond to written communications, regardless of whether the recipient is made aware of the fact, as they don’t get the opportunity to consent.
Media and political engagement advice you can trust from PA Cooperative
Whether you’re in need of support with press releases, reputation management, or either aspects of your company’s communications, it can be tempting to speed things up with the help of artificial intelligence. However, the world of AI is constantly evolving, and so the question of ethics is continually changing, too.
So while AI can be helpful in many ways, it’s wise to seek advice from an experienced public affairs agency to ensure you’re handling your PR ethically and wisely. At PA Cooperative, our experienced team understands the ever-changing business, media and political landscape like nobody else.
We work with organisations large and small to adapt to these landscapes and navigate those evolutions smoothly. Interested in how we can help your business to communicate? Feel free to get in touch to discuss your PR and media relations needs.
