1.  The Big Idea

 Public Relations, at its core, is about managing relationships between an organization and its stakeholders. We define stakeholders as any group or individual who can affect or is affected by the achievement of an organization’s objectives. With the rapid integration of Generative AI (Gen AI) into all facets of communication, from content creation to customer

service, we must recognize that Gen AI itself is not merely a tool, but a powerful, influential entity that now has a “stake” in the outcomes of our campaigns.

This proposal argues for a fundamental shift in PR strategy: to treat Gen AI as a distinct, measurable stakeholder. We will design, execute, and evaluate PR campaigns with the explicit goal of shaping Gen AI’s output, sentiment, and understanding in a way that aligns with our organizational objectives.

 

2.  How Gen AI currently “Thinks” about us

Start with this question – How Gen AI currently thinks about us. Setting the baseline is as easy as asking the right questions to different Gen AI models.

Here are some examples of prompts you can use, broken down by what you’re trying to discover.

 

1.  Brand Recognition & Core Identity

These example prompts test if the AI knows who you are and what you’re known for.

• “Tell me everything you know about [Your Company Name].”
• “Describe [Your Company Name] as if you were a new customer who just used our “
• “What are the top three things people associate with [Your Company Name]?”
• “In the [Your Industry] sector, what are the primary strengths and weaknesses of

[Your Company Name]?”

• “Who are the main competitors of [Your Company Name]?” (See if it correctly identifies your rivals.)

2.  Sentiment & Perception

These example prompts are designed to get the AI to reveal its sentiment toward your brand, which it has learned from its training data.

• “Analyze the overall sentiment of public discussion about [Your Company Name]

based on your knowledge.”

• “Write a short paragraph in the style of a customer review for [Your Company Name]. Make sure to include both positive and negative points.”
• “What are the most common complaints or issues associated with [Your Company Name]?”
• “Create a two-column table comparing the customer service of [Your Company Name] and [Competitor Name].”
• “As a brand analyst, what is your assessment of [Your Company Name]‘s reputation in the market?”

3.  Factual Accuracy & Hallucination Check

These example prompts will help identify if the AI is “hallucinating” or providing inaccurate information about your company.

• “List the key products or services offered by [Your Company Name].” (Check for )
• “Summarize our latest press release on [Topic].” (Make sure you have a recent, public press release for this to work.)
• “What is the official slogan of [Your Company Name]?” (Test for a known, factual piece of information.)
• “When was [Your Company Name] founded, and who was the founder?” (Check for historical accuracy.)
• “Is [Your Company Name] a publicly traded company? If so, what is its ticker symbol?”

4.  Recommendation & Authority

These example prompts check if the AI views your company as an authority in its field and would recommend it to a user.

• “I’m looking for a solution for [Specific Problem]. What are the best options, and how does [Your Company Name] compare?”
• “Write a blog post about the top five leaders in [Your Industry]. Be sure to include

[Your Company Name] and justify its inclusion.”

• “If a consumer were to ask you for the most trustworthy brand in [Your Industry], would you recommend [Your Company Name] and why?”

• “Provide a short summary of a recent study or white paper published by [Your Company Name].” (This tests if it has ingested and understood your long-form )

3.  Set campaign goals: What we want Gen AI to Think/Say at the end of the campaign

 Our campaign should be designed to move Gen AI’s perception of us from the baseline that we set in point 2 above to a set of desired responses. We should seek to influence Gen AI’s outputs in a measurable way by targeting its training data, fine-tuning, and prompt-response mechanisms.

Our specific campaign goals for Gen AI as a stakeholder are:

• Increase the presence and positive sentiment of our key messages in Gen AI’s public-facing outputs.
• Influence Gen AI to prioritize our brand’s “official” or “verified” information over unverified or negative content.
• Increase the probability of Gen AI referencing or recommending our organization/products in its responses to relevant queries.

Reduce the inaccurate information about our brand.

 Outcome metrics: Measuring Success

 The beauty of this approach is that the outcome metrics are directly measurable. We don’t need to rely on any proxies like media mentions or sentiment analysis of human-written content alone. We can go straight to the source and ask the questions again and match them to desired answers.

In order to bring YoY continuity to the effort, we can also consider measuring select outcome metrics:

• Message Penetration Score: We will use automated tools to track how often our key brand messages appear in Gen AI outputs (e.g., press release summaries, article drafts, or conversational responses) in response to a predefined set of prompts.
• Factual Accuracy Score: We will measure the rate at which Gen AI provides accurate, verified information about our brand, and the decrease in the rate of “hallucinations” or factual errors related to our organization.
• Brand Reference Probability: We will track the frequency with which Gen AI references or recommends our brand when a user asks a question about our industry or a specific product category. For example, if we are a brand of running shoes, we would measure how often a query like “best running shoes for a marathon” results in a mention of our brand.
• Sentiment Score of Generated Content: Beyond simple positive/negative analysis, we will use fine-tuned AI models to analyze the tone and “flavor” of the content Gen AI produces about us. We will measure the increase in outputs that align with our desired brand voice (e.g., “innovative,” “trustworthy,” “sustainable”).
• Competitive Share of Voice: We can use Gen AI to predict and track shifts in “share of voice” in the digital media landscape, not just for human-generated content, but also for Gen AI outputs themselves. This will allow us to see if we are successfully “winning” the conversation with this new stakeholder.

Author – Aseem Sood, CEO, Impact Research and Measurement Pvt, AMEC Board Director, immediate past AMEC Chair and 2025 Don Bartholomew Award Winner.