Marketers sound the alarm on AI’s potential to make mistakes and cost firms money

The data: Marketers are voicing concerns about the risks their businesses face when using AI-generated information, per a survey from GWI provided to EMARKETER.

  • 80% of US marketers say their companies have made a serious business decision based on AI-generated insights that turned out to be incorrect or misleading.
  • 45% say their firms took a financial hit after following inaccurate AI-generated information.

Two issues are of particular concern: a lack of verification and outdated information.

  • Just 4 in 10 verify AI-generated insights “most of the time,” and one-quarter do so “occasionally.”
  • 78% say they've used AI-generated information that they later realized was outdated.

Why it matters: As AI integrations become more established, the risks that the technology poses for marketing teams can be more severe.

AI is being used as a core lever for content production, insight generation, campaign execution, and more. That widespread use means AI errors can have real consequences, and reputational and financial damage from those errors can have long-term effects.

Marketing teams carry outsized responsibility because the work they do reaches far beyond the internal organization. Yet they’re facing AI readiness problems—45% struggle with the quality of data and 39% have difficulties keeping AI compliant with brand guidelines, per Epsilon.

If teams don’t step up AI output verification, they will continue to expose their brands to serious risks.

Recommendations for marketers: Refocus AI implementation on responsible governance, even at the expense of adoption speed.

  • Prioritize verification over adopting new AI tools. Consider establishing an adoption pause until teams report higher rates of verification.
  • Run secure experiments for every AI tool in play—these tools are constantly updating and therefore require ongoing monitoring.
  • Work with internal engineers or external partners to map out exactly how AI tools connect to each other within company ecosystems. Knowing these connections will help trace errors to a precise source.

You've read 0 of 2 free articles this month.

Get more articles - create your free account today!