Image Repair Differences Between Human and AI-Generated Messaging: The Case of the Chicago White Sox

With 121 losses in 2024, the Chicago White Sox set an unwanted record for the most losses in a single Major League Baseball season. Team owner Jerry Reinsdorf wrote a letter apologizing to fans following the season, including additional attempts at image repair. The increasing availability of artificial intelligence (AI) raises unique questions concerning how apology messages ought to be generated and how effective such messages are with disgruntled fans. To determine if the real message from Reinsdorf was effective, we generated a similar apology using AI. Fans were randomly assigned to read the real message from Reinsdorf, one generated by AI, or a no message/control condition. Results of this experiment did not reveal significant differences across the three conditions – in fan perceptions of organizational reputation, intention to purchase season tickets, or intentions to subscribe to cable and streaming services carrying the games. Follow-up social media analysis indicated strong negative sentiment directed specifically at Reinsdorf, which abets the claim that a tainted communicator may not be effective regardless of how well the message is crafted, making the lack of significant differences an important finding. Discussion of the results, given the purpose of Reinsdorf’s letter, applies image restoration theory to explain the case, and we offer a new method for resolving social media sentiment analysis ambiguity.