AI usage increases daily. In fact, a report from the Federal Reserve Bank of St. Louis says workers spend roughly a quarter of their business day using intelligent automation to complete tasks. Looking more closely at agencies, 75% of PR and marketing pros are turning to AI for everything from campaign brainstorming to developing content.
But let’s face it: Automating workflows can pose major operational and security challenges as well as reputational risks without proper guardrails. And yet, only 50% of PR and marketing firms have a clear set of restrictions outlining the ways AI can (and can’t) be used. Agencies that fail to lead on governance will quickly fall behind those that do. Merritt Group rolled out a formal policy around AI utilization more than a year ago. However, we know that our standards must adapt as AI evolves and matures.
Our in-house AI task force, MG Labs, recently conducted an extensive audit of the AI policies produced by other mid-sized and global communications firms. We analyzed these findings to not only understand the industry’s broader POV on AI use, but also to crystallize the updates to our own policy. Here’s what we found.
From experimentation to accountability
Agencies can no longer afford to view AI adoption as a “test and learn” exercise without providing more structure for use and adoption. Although most agency standards are internal-facing only, several global agencies have publicly defined what it means to use AI securely and responsibly:
- Clear boundaries: First and foremost, follow the AI standards of your clients. For example, if a client has prohibited their own employees from using generative AI tools, it’s important to adhere to the same guidelines on the account team side. If clients have less restrictive (or undefined) policies, err on the side of caution when inserting personally identifiable information (PII) or sensitive client data into AI tools.
- Human accountability: Never treat AI-generated content as a finished draft. Instead, use AI-generated content as an initial framework that you must review, edit, and fact-check before it’s published online or sent to a client.
- Cross-functional oversight: Create AI standards in collaboration with ethics, DEI, digital, and legal teams.
These are not just compliance measures—they’re trust-building mechanisms. In marketing and PR, trust is currency.
Test, build, and promote
Communications agencies of all shapes and sizes are putting more effort behind AI testing and exploration. Similar to Merritt Group, we are seeing a handful of agencies create their own user groups that safely assess emerging AI tools.
Some global firms are taking their experimentation with smart technology a step further by customizing large learning models (LLMs) for their proprietary market research efforts. These LLMs discover, evaluate, and synthesize the most common factors that impact societal trust with greater precision and predictability.
As arbiters of truthful information, an increasing number of PR and marketing executives are also penning monthly op-eds and blog posts to raise awareness of the quality AI tools that industry professionals should use in their daily workflows. It’s clear that the strongest agencies understand AI’s value and are taking a myriad of proactive measures toward responsible use.
The next frontier isn’t adoption—it’s governance. Agencies that lead here will differentiate themselves not just as innovators but also as responsible stewards of client trust in the AI era.
Want to learn more about the latest developments in artificial intelligence? Visit the MG Labs page to explore our AI tool reviews, testing insights, and more.
