The Real Work of AI Integration
Building Ethical, Durable Systems for Creative Teams
This essay was contributed by Matt Eaton and Hope McGowan , founders of Nepholve. Helping media organizations accelerate digital transformation.
The rapid evolution of generative Artificial Intelligence (AI) models has fundamentally reshaped the content supply chain, shifting the focus from linear production to accelerated iteration and deployment. The second half of 2025 has seen the pace of technological advancement increasing significantly, bringing forth new capabilities such as Sora 2 for video generation, Ray 3 offering the first HDR content, and Google VEO 3.1 enabling enhanced text-to-video and storyboarding. There is a danger, however, of focusing too much on the technology, rather than the outcomes and how these AI capabilities are operationalized.
Forbes recently covered an example of McDonald’s pulling its 45-second spot. It provides a warning to other brand creative teams. Consumers perceived the result as “AI slop” with rapidly cut scenes provoking feelings of horror and anger, rather than empathy. Far from being a video created in a blink of an eye, the creative team behind the controversial ad initially complained that they had spent weeks of sleepless nights perfecting the prompts to generate the content. This points to capabilities that many creative organizations are missing in their management of a more iterative production cycle, such as tracking the lineage of content and the AI prompts used to generate components used in the final production.
Whilst AI allows us to visualize ideas quicker, ultimately, to be successful, creative teams still rely on humans:
Human judgment to determine which tools and treatments should be used to get the best results. AI lacks the ability to understand context and predict how consumers are likely to respond to the content.
Human craft to bring creative ideas, tell a story that will resonate, and push the envelope. AI is based on patterns from the past and its results may be too safe or corny on its own. It needs a human to create the next BIG idea.
Underpinning these are two important areas often overlooked by creative teams; an ethical framework and a more agile, cross-functional operating model better able to manage an iterative production process.
The Foundation: Ethical Frameworks for AI Governance
Creative organizations need a strong ethical framework for AI which addresses technical and operational governance across the entire content lifecycle. This needs to be tailored to the individual organization, but here are some of the most important tenets to consider.
Accountability is key to any ethical framework, requiring organizations to take full responsibility for the outcomes produced by generative AI. This responsibility is best managed through specific governance structures, ethical guidelines, safe prompt usage, and continuous monitoring of AI models available in the market. Furthermore, it is paramount that generative AI serves to support and amplify human creativity, rather than substitute the creative process.
The provenance and models used by the chosen generative AI platforms must be understood from both a technical and an editorial perspective. This requires knowledge of the training data and the inherent biases in some models to ensure responsible use. Importantly, this also means that copyright and clearances needed for components used in the content are recognized and tracked.
Transparency is also a key tenet of an ethical framework. Organizations must be transparent not only with consumers, but also staff and the wider community, regarding how generative AI has been utilized to produce the final content. Linked to this, an ethical framework should foster a safe psychological space, encouraging internal personnel to speak up about any concerns they might have regarding the use of generative AI.
Perhaps the most important tenet is that the framework should be recognized as work in progress and regularly reviewed, as the underlying technology is constantly evolving.
Reimagining the Operating Model: The Creative Squad Structure
The use of AI toolsets changes the way creative projects run. They encourage a test-and-learn approach to content creation, enabling rapid development of creative treatments for review. Feedback, often collected from sample consumers, can then be incorporated into the creative process. In this way, content can go through multiple iterations, with tweaks to an element here or an element replaced there. Ultimately, the workflow from ideation to delivery and activation becomes less linear. AI tools also change the skill sets needed on a creative team by providing a wider range of treatments, from 3D modelling to animation. This means a broader range of human creative skills and ways of working is required to manage the different aspects of a production. The need to harness an array of skills and the iterative nature of the creative process require a new operating model.
An iterative, cross-functional way of working is common within software development teams, and a similar squad idea may be applied to creative teams. At the heart of this new operating model is the goal of ensuring AI tools consistently amplify human creativity rather than replace it.
Figure 1 Cross-functional Creative Squad Team Structure
Creative squads operate across the content supply chain stages: Ideate, Generate, Create, Adapt, Deliver, and Activate. As content moves through these stages (e.g. from copy generation and storyboarding in Ideate and Generate, to localization and personalization in Adapt, and finally to predictive performance in Activate), the squads receive real-time feedback. This allows the team to iterate output quickly, making element tweaks or replacements, demonstrating the flexibility AI tools provide.
At the core of the Creative Squad is the Brand Producer, who maintains consistency, direction, and messaging across the project. The squad may include specialized roles:
Creative Coordinator: Functions akin to a project manager or Scrum Master, managing the workflow.
Editor: Focuses on content refinement and assembly.
Technology Expert: Scans the market to deploy the optimal AI models and technical tools for the project.
Brand Experience Expert: Ensures creative output aligns with the consumer journey.
Market Research & Analytics/Performance Specialists: Provide real-time data feedback on creative output.
A single creative squad cannot deliver value in isolation; there needs to be an operating model that supports this new way of working. Like ethical frameworks, operating models vary depending on the organization’s strategic priorities and leadership style, but the diagram below illustrates how creative squads might be embedded within a broader context.
Figure 2 Creative Squads within a Broader Operating Model
Starting at the top of the diagram, an overall brand management function looks after budgeting, prioritization, and strategic direction. The change management component of this function is critical to ensuring teams adopt the right ways of working. Projects are scoped and prioritized before being fed into a demand hopper for assignment to the creative squads, providing a structured approach to project initiation.
There may be one or more creative squads composed of cross-functional team members working on the projects. Creative Loops allow for feedback to be received from consumers in real-time on the performance of their squad’s output and they iterate the content, if required. The iterative value drops are measured along with the benefits to ensure the squads are delivering to plan.
An Enterprise AI Capabilities team supports the operating model, providing the necessary infrastructure and governance for the creative squads to operate effectively. This centralized capability is responsible for:
Tool Provisioning: Ensuring squads have the appropriate tools, platforms, and skills to deliver projects.
AI Governance: Upholding ethical and technical standards.
Data Security and Management: Overseeing the security and handling of training data and generated assets.
New Technology Onboarding: Scanning for and integrating new models while removing operational blockers.
Ultimately, the focus must remain on the outcomes, ensuring that the adopted technology amplifies human creativity, increases content velocity, and delivers more successful campaigns, rather than being distracted solely by the novelty of the technology itself.
As AI technology continues to mature, organizations have an opportunity to build durable, ethical systems of working centered around creative teams. To avoid creating ‘AI slop’, human judgment and craft should remain central to the creative process. First, define your organization’s ethical framework for AI and consider how it will be used across the entire content lifecycle both from the perspective of employees and end-consumers. Changing ways of working, team structures, and operating models takes time and a champion within executive leadership. There may be skillsets that need reinforcing and will likely involve adopting new ways of measuring and rewarding performance.
Start the journey towards creating an inter-disciplinary creative squad with a proof of concept focusing on a single project or brand. Establish new iterative working rituals, like daily stand-ups, shared messaging forums, and common tooling. Consider how best to integrate feedback loops into the creative process to improve content over time. The role of the Brand Producer is critical in these early stages, building squad cohesion and encouraging new ways of working. Share ideas, learn, and implement new improvements that work for your organization. Over time, additional cross-functional creative squads can be created and adopt similar ways of working. Introduce supporting functions into the operating model, such as the Enterprise AI Capability, that are focused on removing blockers for creative teams.
 | A guest post by
|