Recently released posters promoting the feature film Civil War stirred controversy. But not the kind of controversy that its studio parent A24 was looking for. The AI-generated visuals used in the posters caused immediate backlash among commercial artists and consumers over their inaccuracy and lack of production quality. One depicts a military gun boat on Los Angeles’ Echo Park lake next to a random giant swan. Another poster places the Chicago Marina Towers on the wrong side of the Chicago River. Both are widely regarded as generative AI fails.
AI creative pollution proliferates
Sadly, AI fails like Civil Wars’ are symptomatic of a larger issue of AI Creative Pollution, defined as “average or offensive AI-generated commercial creativity.” Havas Chief Strategy and Innovation Officer Mark Sinnock told us, “…in their rush to elevate the middle and create a glossy universal average, they’re also creating a culture of blandness and desensitization.” From the Taylor Swift deepfake scam ads to Google Gemini’s historically inaccurate images of German soldiers to Netflix’s distorted images in its true crime documentary, What Jennifer Did, consumers cringe at the uncanny valleys of AI fails polluting the internet.
Wracking the creative nerve
Whether advertising or entertainment, AI creativity causes considerable anxiety and dissonance among commercial artists. The use of AI-generated creative comes on the heels of a hard-fought contract negation between the Hollywood studios and the Writers Guild of America (WGA) in which WGA came away with significant concessions about the use of AI in film production. AI pollution symbolically challenges that win. It also touches a nerve in broader American society where a third of US consumers (who have heard of it) distrust generative AI. And it strikes fear in the hearts of employees in creative industries who perceive AI as a job threat, despite a deep embrace of AI technologies and tools. The reality is that by 2030 automation and generative AI will replace 7.5% of U.S. advertising jobs.
“Give a hoot and don’t pollute”
To clean up all this AI creative pollution, the industry needs a marketing operating system that leverages the economic efficiencies of AI to achieve superior performing creative quality. We call this concept “brand AI models,” which Forrester defines as brand-specific, AI-powered marketing operating systems comprising genAI foundation models, machine learning, customer insights, media data, brand IP, and human expertise that produce personalized marketing campaigns and customer experiences.
Brand AI models are not replacements for creativity. They combine the efficiency, scale, and precision of technology with the artistry, intuition, and tastes of artists. Most importantly, they are trained with your company’s brand IP — the tangible and intangible assets that distinguish your company’s goods or services from that of your competitors — using data sets like product IP, brand essence, logos, fonts, colors, tone of voice, images, previous advertising, and/or brand idea platforms.
Reinventing how business value is created
When AI-generated creative becomes equally cost efficient and effective, brands and service providers see multiple advantages:
- Cost savings through automating repetitive tasks.
- Increased productivity by enhancing the workforce with tools that speed up and improve their work.
- Improved quality of outputs based upon machine learning optimization — as your algorithm learns and improves, the average and awful AI pollution turns into uniqueness and relevant creative that touches the spirit and emotions of its audience.
- Eventually, brand AI models produce net new growth by understanding the unmet needs or pain points of customers and recommending new products and solutions to bring more value.
Originally posted on Forrester
