Organisations have traditionally reacted to a crisis only when it’s gained enough traction to hove into view. They then rely on a combination of processes, experience and gut instinct to respond. So what can AI do to help them?
AI has had the most impact on PR and crisis-handling through media monitoring. Machine learning techniques mean that AI-powered monitoring tools can scan thousands of online sources (in particular) and identify key terms and trends that could signal the beginning of a crisis (as long as it knows what it is looking for). Predictive analytics can also help PROs anticipate crises by analysing historical data and current trends to identify potential vulnerabilities.
Then during a crisis, AI can monitor the potentially hundreds of thousands of online and social media interactions in a way that humans cannot, allowing for a more rapid and possibly tailored PR response. Dashboards can generate, and PR teams can rapidly disseminate, data and information on which crisis management teams can make decisions (as long as the humans are willing to trust the data).
Furthermore, Chat GPT and Gemini etc can of course generate statements and press releases much more quickly than humans; maybe it can help with responding to customer inquiries (although remember how KFC ran into trouble in November 2022 with its AI automated marketing message urging Germans to ‘commemorate’ the 1938 Nazi Kristallnacht pogrom of Jews with cheesy chicken (see our case study at the time here).
This helps during a crisis when speed of response as well as accuracy can be key to getting on top of and ahead of the issue. That said, it should of course be remembered that audiences look for emotional intelligence and sincerity from brands, so human oversight of AI-generated responses (which we all know are currently quite bland) is crucial. The right tone of voice and emotionally engaging and compelling messaging are unlikely to be generated by AI as it currently stands.
Thus, the reality of integrating today’s AI into crisis prediction and response is that yes, it can crunch and deliver data swiftly to anticipate and aid crisis decision-making. But it still needs humans to think about angles, messages and tone, decide which stakeholders to speak to and when, who to put forward as a spokesperson, and what and how they should communicate.
You cannot always rely on models like ChatGPT. It doesn’t access a database of truths; it merely suggests answers based upon patterns from the data on which it was trained. A PR team can look at a crisis and know that, for example, their COO faced a similar situation at a previous company and handled it brilliantly; they also know when the CEO should not be the face of the organisation (BP’s Tony Hayward, anyone?). Chat GPT cannot yet replace detailed experience, knowledge and judgment.
Without a doubt, PR teams can leverage AI to assist their crisis preparation and response. If a brand/ company is big enough and invested enough in AI monitoring and analytics, then of course this can all be tailored to the organisation which will make it more useful. However, how many organisations have the resources to do so?
Effective crisis management is the result of diligent preparation, planning, training and practice. Which is why Electric Airwaves’ crisis planning and preparation services, crisis media training and crisis exercises are still in demand. Whatever the scenario, successful handling of a crisis comes down to people, processes and plans. We can generate scenarios that have AI as part of them – for good or for ill – but the confidence to handle crisis comes from experience. Experience comes from learning. A crisis is not the time to start learning.