Georg von Richthofen, Sonja Köhne, Maja Golf-Papez
9 min
Generative AI (GenAI) is fundamentally altering creative work, yet the professional response within fields like advertising is far from uniform. Some creatives embrace these tools as catalysts for productivity, while others view them as existential threats to their livelihoods. This paper uses the lens of technological frames—the shared assumptions and expectations a group holds about a technology—to explain why creatives respond so differently to GenAI.
The authors identify that creatives interpret GenAI through three primary frame domains:
These interpretations drive distinct, actionable responses among professionals. When creatives view GenAI as a tool or collaborator, they typically engage in AI skilling, integrating the technology into their current workflows. When they perceive GenAI as a disruptive agent, they may pursue reskilling to transition into less-automated roles (such as strategy or account management). Finally, those concerned with the homogenization of creative work often choose deep skilling, focusing intensely on human-centric capabilities—like cultural sensitivity, creative judgment, and empathy—that AI cannot easily replicate.
This research moves beyond simple dichotomies of AI acceptance or resistance. It provides managers with a structured way to understand the underlying anxieties and expectations of their creative teams. By recognizing that these responses are tied to specific mental frames, leaders can better tailor their communication, redesign workflows to keep humans in the loop, and develop capability-building initiatives that reflect the diverse needs of their workforce.
Generative artificial intelligence (GenAI) is transforming creative work, yet creatives do not respond uniformly. To explain this variation, we draw on the concept of technological frames, three years of netnographic research in online advertising communities, and interviews with advertising creatives. Our analysis identifies competing interpretations of GenAI—regarding its level of agency in the creative process, its impact on jobs, and its role in shifting how value is created in outputs—that converge in distinct narratives about its implications for creative work. These narratives, in turn, drive distinct adaptive responses among creatives: AI skilling, reskilling, and deep skilling—the latter involving deliberate cultivation of human capabilities that GenAI cannot readily replicate. The findings inform both research on GenAI in creative work and managers navigating its adoption.
Alex: That makes the discussion useful, but leaves uncertainty about who is speaking. Did they check the patterns against identified professionals?
Sam: They supplemented the forum analysis with interviews involving twelve advertising creatives. Participants included art directors, creative directors, and a copywriter, working across different agency sizes. Most interviews followed initial forum coding, allowing the researchers to challenge and refine emerging interpretations.
Alex: Then the method’s central move is linking what people say AI can do with what they say they’ll do.
Sam: Yes. One forum contributor said clients often accept AI copy that is good enough and much cheaper. That contributor then described diversifying into strategy work. The authors treated this as a connection between expected output value, threatened jobs, and a career response.
Alex: Let’s unpack those interpretations. What changes when someone sees AI as more than a tool?
Sam: The authors distinguish a tool, a collaborator, and an agent. A tool supports a task under human command, like suggesting alternative wording. A collaborator participates in an iterative exchange, helping someone explore and revise ideas.
Alex: The collaborator description still leaves a person making creative decisions. What does the agent description involve here?
Sam: It shifts more of the creative work to AI, while humans set goals and potentially check outputs. One interviewed director described an agency system with AI roles handling strategy, concept review, and production. A client briefing went in, and finished assets came out.
Alex: That sounds consequential, but is it representative of how these creatives were working?
Sam: The agent frame was relatively rare, though the authors noticed it more in recent discussions. These are interpretations and reported practices, not a measured distribution of agency workflows. They also describe someone initially treating AI as a collaborator, then downgrading it to a tool after disappointing results.
Alex: So these are not permanent personality types. Experience can change how someone interprets the same technology.
Sam: The authors describe a community where competing frames coexist and evolve. On jobs, some anticipate disruption, including displacement. Others expect gradual evolution: less creating from scratch, more editing, curating, directing, and maintaining client relationships.
Alex: Is disruption always the pessimistic view? Faster creative work could also open possibilities for smaller teams.
Sam: Disruption includes optimistic interpretations too. Some contributors anticipate resourceful small teams producing much more. The authors also identify a narrative where AI as a collaborator enhances creativity, productivity, and enjoyment through rapid experimentation.
Alex: There’s another issue besides who does the work: whether the resulting advertising becomes interchangeable. How does that enter the argument?
Sam: Some creatives expect widespread AI use to make brands sound and look similar. They connect that to clients prioritizing efficiency and lower costs. Others see the same homogenization creating demand for originality, authenticity, and visibly human work.
Alex: That second view is an appealing market prediction. Did the authors test whether clients actually pay more for human originality?
Sam: No. It is an expectation expressed by creatives, not an observed market outcome in this study. The contribution is showing how that expectation supports a particular adaptation strategy: investing more deeply in human creative capabilities.
Alex: Bring the career responses together. What does each strategy ask a creative to learn?
Sam: AI skilling means learning to use the technology effectively, through courses, tutorials, experimentation, or colleagues. Reskilling means developing capabilities for another role, such as strategy or human resources. Deep skilling means strengthening craft, creative judgment, empathy, and cultural understanding.
Alex: Deep skilling sounds more specific than refusing the technology. It’s an argument about where distinctive value might remain.
Sam: The authors present it as a deliberate investment in human expertise. Examples include practicing writing for different audiences, studying their subcultures, and creating without AI to train judgment. It differs from technical fluency, but the study does not establish which pathway is most effective.
Alex: Are the links between beliefs and strategies clean? Someone could see AI as a tool and still expect fewer jobs.
Sam: The authors explicitly give that example. The relationships are recurring patterns, not mechanical rules. Reskilling can also appeal to someone expecting job evolution rather than elimination, if the altered work seems less enjoyable or valuable.
Alex: With thousands of posts, what is the biggest reason not to treat this as a prediction of creative work?
Sam: The data cannot establish how prevalent these frames are or test the proposed relationships. The interviews showed no salient discrepancies with the forum analysis, which supports the framework. But agreement across sources does not demonstrate causation or career protection.
Alex: And a senior campaign writer may face a different situation from a junior producing routine content.
Sam: Participants repeatedly made those distinctions. They expected specialization, seniority, skill, and clients’ priorities to matter. Those are contextual expectations in the data, not tested guarantees that strategic or senior roles are safe.
Alex: Who should read the full document, and where should they start?
Sam: Researchers studying AI adoption, creative work, or career adaptation should start with the section on interrelations and contextual influences. It shows how the interpretations connect to strategies. Then read the limitations, especially the unanswered questions about prevalence and effectiveness. Managers should also read the managerial implications for differentiated training and workflow design.
Alex: And what should everyone else carry away?
Sam: People adapt to AI based not only on what it can do, but what they think will remain valuable.
Alex: Thanks for listening.