3 Ways GenAI Can Help With Issues Preparedness

May 2024
Share this article

You’ve probably heard the quote, “A lie gets halfway around the world before the truth has a chance to get its pants on.” 

In today’s digital whirlwind, with social media turbocharging falsehoods and deepfakes distorting reality, the truth is struggling even more to keep pace. 

The velocity at which information spreads demands an immediate crisis/issue response, achievable only with thorough readiness. Preparing for potential issues is labor intensive, but Generative AI (GenAI) can help with the heavy lifting from instant horizon scanning to drafting strategies in minutes. 

Use Perplexity.ai for horizon scanning.

Horizon scanning is all about learning from past communication or reputation hiccups and thinking ahead to what might come up. GenAI can expedite this process by promptly generating a comprehensive list of issues for deeper examination.

The best tool for this exercise is Perplexity AI, which leverages Open AI’s ChatGPT models, Anthropic’s Claude 2, and Google’s Gemini to search the Web and compile information from multiple sources. What’s helpful is that Perplexity links each statement to clickable sources for verification. All you need to do is put in a specific query or prompt.

While naming a product in the prompt will generate more specific and valuable results, let’s use a redacted example (“U.S. weight loss drug”) just in case confidentiality is a concern:

Prompt: “Provide a list of concerns about weight loss drugs published by U.S. news outlets in 2024.”

Perplexity AI answer: 

  • DA investigations into potential risks
  •  Suicidal ideation
  • Hair loss
  • Food accidentally inhaled into airways
  • Supply issues

Perplexity elaborates on each bullet with an AI-generated summary and clickable sources. In this case, the sources were U.S. News & World Report, Healio, Reuters, and CNBC. 

Prompt: “Provide a list of emerging concerns weight loss drugs will likely face in 2024.”

Perplexity AI answer: 

  • Supply shortages
  • Off-label prescriptions
  • Regulator scrutiny relating to rare side effects
  •  Inequitable access to treatment
  • Affordability
  • Insurance coverage
  • Long-term efficacy

While GenAI provides an excellent starting point, it is crucial to integrate human insights from a team intimately familiar with the product or service. This ensures compiling a comprehensive list encompassing nuanced scenarios, particularly if using redacted information in the prompts.

Use ChatGPT or Gemini to stratify risk.

Once you have a complete list of potential issues/crises, ChatGPT 4 (paid version) or Gemini (free version) can help stratify risk and identify high-priority scenarios that are both highly likely and high impact. Here’s how GenAI would be prompted for the weight loss drug example:

Prompt: “Using the attached risk assessment grid, score the following potential issues for a U.S. weight loss drug: 

  • Supply shortages
  • Off-label prescriptions
  • Regulator scrutiny relating to rare side effects
  • Inequitable access to treatment
  • Affordability
  • Insurance coverage
  • Long-term efficacy

Here’s the risk assessment grid to attach. Take a photo or screenshot and attach it to your prompt.

Risk Assessment Grid

While ChatGPT 4 generates scores for each issue in the order they were presented, Gemini takes a proactive approach by calculating an overall risk score and delivering a prioritized list of top-quadrant scenarios to focus on first. If you want ChatGPT 4 to do the same, use this follow-up prompt:

“Calculate the overall risk score for each issue by multiplying the likelihood score by the impact score for each issue. The higher the score, the greater the overall risk. Prioritize the list of issues by overall risk score under the risk categories in the risk assessment grid: low-priority, midway scenarios and top-quadrant scenarios.”

Use ChatGPT or Gemini to brainstorm risk-mitigation strategies.

Whether the plan is to prepare for all potential issues or focus on top-quadrant scenarios, ChatGPT and Gemini can generate PR-focused risk mitigation strategies. Here’s the prompt: 

“Provide a list of public relations-focused risk mitigation strategies for [insert product or product category] facing the following issues in the [insert country/region] market: [insert list of issues].” 

The more specific the inputs, the more actionable the strategies will be. However, even if using redacted information, GenAI is still a fantastic ideation partner by generating a useful draft that can be refined in collaboration with the tool or by a human team.

Navigate confidentiality constraints.

A significant challenge in leveraging GenAI for issue preparedness is handling confidential information. Understanding why inserting sensitive data into AI tools can be problematic is essential. 

These platforms often store and process data on external servers, raising concerns about privacy and security. This poses a real challenge for PR professionals bound by nondisclosure agreements (NDAs) and organizational AI guidelines. 

Familiarize yourself with the ethical guidelines and compliance standards of the AI tools you’re using. Some platforms offer data privacy and confidentiality options that may align with NDAs or an organization’s AI policies. 

Instead of clear guidelines, connect with legal and IT before integrating GenAI into your workflow to establish safeguards for using the tools responsibly.

Always have a human in the loop to review AI-generated content for inaccuracies, fact-check data/sources, and prevent the unintentional disclosure of sensitive information.

Here are three ways to outsmart confidentiality constraints:

  1. Redact sensitive details. When using AI to generate content or strategies, ensure that all confidential information is either omitted or adequately masked. In this context, that also may include masking the company, product name, or any other identifying details. For example, in this article, “a U.S. weight loss drug” is used in lieu of a specific drug in the GenAI prompt examples.
  2. Substitute data with placeholders. Utilize dummy data to represent sensitive information to benefit from AI’s analytical prowess without risking exposure. 
  3. Champion a secure environment. Advocate for integrating AI tools within a protected infrastructure, such as behind a company’s firewall, to safely explore its capabilities. Depending on the sector and whether the confidential information is market-moving, there may still be varying levels of confidentiality constraints.

These use cases offer a glimpse into how GenAI can boost issue preparedness, paving the way for deeper dives into GenAI’s many applications in public relations.

Return to Current Issue Crisis Management | May2024
Share this article

Subscribe to Strategies & Tactics


*Strategies & Tactics is included with a PRSA membership