Generative AI

WASHINGTON — Generative AI remains at the center of conversations in the worlds of technology and advertising, with advancements continuing to roll out from tech giants, agency holding companies and marketing consultancies. While causing some uncertainty about job futures, the technology is also seen by marketers as a balm for burnout and a way to boost investment in creator content.

Amid all of these applications and experiments, generative AI still faces a raft of legal issues and practical pitfalls that marketers must navigate while integrating the tech into their operations. Those concerns were the topic of a panel at the IAB Public Policy and Legal Summit in April.

Panelists also clarified some definitional distinctions that marketers must understand, especially as agencies, ad-tech providers and other platforms rush to adopt generative AI while rebranding and highlighting AI functionality that has been part of the ad industry for more than a decade. 

“You've probably been using machine learning and deep learning to segment your audience, to develop ad budgets, to place ads, to understand what type of viewers may be more responsive to particular types of advertisements,” said Dera Nevin, managing director at FTI Consulting. “Machine learning and deep learning has been used in the advertising industry for a long time … and now we're starting to see the use of generative AI to generate content.”

Cooking with AI

To understand AI, Nevin suggested a controlling metaphor around cooking wherein algorithms are recipes, data inputs are ingredients and generated outputs are prepared food. While machine learning is a simple recipe, the deep learning that drives large-language models and generative AI is a much more complicated one. As in the kitchen, the final product is only as good as the ingredients, and the data which AI is trained on limits the efficacy and accuracy of the output.

“In order to really understand what kind of food you're going to get when the recipes interact with the ingredients, you actually need to know what's in the kitchen and who's preparing it,” Nevin said. “But there's often little transparency behind what the [recipe] is or what the ingredients are. Without knowing that, you just don't know what kind of food … is going to come out.”

Agencies and brands should be concerned about what data they are inputting as generative AI prompts as well as the output that is generated. When using public-facing generative AI tools like ChatGPT, that data becomes part of the algorithm's data set — whether it is confidential, personal or otherwise private.

When considering output, marketers should be wary of acontextual content that is generated when AI doesn't understand context, resulting in output that can be embarrassing to brands. Plus, in its attempts to simulate human behavior, AI can click or “behave” as human beings, driving incorrect metrics or understandings of engagement. 

“Hallucination” has become a popular way to describe the unexpected output generated by AI, but Nevin pushed back on the term as it attributes human characteristics to technology. So-called hallucinations are happening due to underlying math and probabilities; the technology is doing what it is designed to do, but does not have the human ability to create original ideas.

“A human being combined two concepts to come up with ‘Sharknado.' I don't know that an AI could do that," she said. "But an AI could come up with very credible ‘Sharknado 2’, ‘3’, ‘4’ and ‘5.’"

Generating opportunities

Much of the last year has been focused on the threat vectors of generative AI, including concerns around publisher and ad traffic, compensation for the inclusion of copyrighted material in large-language models and signaling protocols for determining what should and should not be allowed to train AI models, explained IAB CEO David Cohen in an interview at the summit. As the industry begins to solve for those threats, brands and agencies can focus on the opportunity vectors.

“How do we use all of this for creative efficiency, workflow efficiency, making our business more agile, adaptable and efficient? There's tons and tons of work that's going on there on the creative side,” Cohen said. “The opportunity [piece] is what the next 12 to 24 months will look like.”

For marketers, some of those opportunities might be better served by looking to tech company offerings that protect brand safety more effectively than public-facing tools. For example, Adobe and its Firefly platform can be trained on a brand's assets — rather than publicly scraped data — giving the outputs better resonance with the brand, explained Matt Savare, a partner at Lowenstein Sandler, LLP, during the panel. Plus, companies like Adobe, Google and Shutterstock have announced plans to indemnify users against third-party intellectual property claims, protecting smaller brands and agencies from legal peril.

Still, those opportunities might have to wait as brands and agencies triage more pressing concerns, like the deprecation of third-party cookies, first-party data strategies and Google's Privacy Sandbox proposals, Cohen explained.

Whether generative AI will eventually deliver on the loftiest promises of its biggest boosters and be able to create high-level, creative campaigns remains to be seen. Other technology has revolutionized so many parts of the advertising ecosystem in previously unbelievable ways, and generative AI could be the next miracle tool — one day.

“The generation that's coming up is going to be doing what we do in extremely novel ways, and I wouldn't be surprised if we get to precision advertising [with AI] at some point," Nevin said. "I just don't know how quickly that's going to come.”

Ninety-one percent of U.S. advertising agencies are either currently using (61%) generative artificial intelligence (AI) or exploring use cases (30%) for the tech, outpacing other groups including marketing organizations and the general business population, per a recent report from Forrester that was shared in June. 

Large agencies are further along the generative AI adoption path, with 78% of respondents at agencies with more than 201 employees saying they leverage the tech compared to only 53% of small agencies, or those with less than 50 employees.  

More than half of respondents expect generative AI to have significant or very significant impact on key aspects of their agency’s ecosystem in the next two years, though concerns remain around legal liabilities, copyright infringement, data privacy and security.

Generative AI has moved from the theoretical to the practical in 2024, according to Forrester’s report, titled “The State of Generative AI Inside U.S. Agencies 2024.” More than three-quarters of respondents said generative AI is a disruption for their agency. Twenty-nine percent of respondents said it is a major disruption that will change their business forever.  

Right now, most of that disruption is happening at creative agencies. Some 69% of respondents at creative agencies reported they currently use the buzzy tech. However, 57% of respondents at media agencies said their agency has begun using generative AI, and that percentage will likely rise as teams figure out new ways to leverage the tech’s speed, insights and innovation.

The latest report from Forrester is derived from a joint survey with the American Association of Advertising Agencies (4A’s) fielded during April and May. Respondents were decision makers with titles of vice president or above at U.S. agencies. 

Respondents identified four key areas of an agency’s ecosystem where generative AI will have significant or very significant impact: how an agency creates content (76%), the agency marketplace (71%), how consumers interact with the work created by an agency (69%) and what content an agency produces for clients (62%).  

One main area in which agency executives are hoping to derive value from AI is productivity. Around of the respondents using (54%) and exploring (46%) generative AI indicated they were looking to the technology to improve the speed and quantity of creative ideation and production. Indeed, 74% of respondents at agencies using the tech said aiding creative ideation and brainstorming is a high or critical priority. Nearly half (49%) said that generating assets for dynamic creative optimization is a high or critical priority. 

Agencies are also using generative AI to summarize audience insights and media reporting, with 59% of respondents using the tool to prioritize organizing consumer audience data sets into insights and 49% using it to summarize marketing performance results. Many are also using it to augment internal workflows and improve productivity, such as creating consumer profiles and personas, analyzing competitors and transcribing and summarizing calls and presentations.  

Despite widespread adoption, barriers remain. Chief among them are legal concerns, such as IP ownership, attribution, copyright infringement and delineation of liability. Ninety-four percent of respondents exploring and 83% of respondents using generative AI indicated these issues were challenges. Other concerns include data privacy and security, governance and risk, reliability, accuracy and bias. Notably, employee resistance is also expected to be a barrier to adoption. 

“This human obstacle stands apart from the others because employee concerns exist entirely within the agency but are irrationally unpredictable and pertain to employees’ emotional safety,” according to the report. “Employee resistance or lack of AI expertise stands ready to undermine agencies’ value, product, and technology investments as well as a decade of agency business model transformation.”

Meta unveiled a set of generative artificial intelligence (AI) features for advertisers in May 2024, including image and text generators that speed up creative production. The offerings address a demand among marketers for a greater sense of control in their AI experiments and a guarantee that the end results will meet what are often specific internal brand guidelines, like ensuring the correct colors are present in ads.

Striking the right balance with generative AI is important for Meta to build confidence in a technology the Instagram and WhatsApp owner believes is key to future growth but that has fallen into complex ethical and legal tangles.

“Part of the answer for how to grow [generative AI] adoption is actually making sure it’s not just a black box but you also have the right levels of control,” said John Hegeman, vice president and head of monetization at Meta, during a Q&A portion of a presentation to reporters. “Brand guidelines are an area where brands are going to continue to have a lot of specific preferences and not want those to be violated.”

Meta’s image generator at launch only spits out image variations based on assets that the advertiser submits rather than working off of text prompts alone. A business may have an existing picture that shows a cup of coffee against sunny farmland and could ask Meta to show the cup instead “surrounded by coffee beans and lush leaves,” leading the software to produce several potential image ideas that might serve as a replacement.

“Most marketers and advertisers prefer utilizing the assets that they’ve provided,” said Alvin Bowles, vice president of Meta’s global business group, during the presentation. “The guesswork, you take some of that out of it because it’s actually coming from the agency and client.”

Meta will eventually let the image generator function off of text prompts, according to Hegeman, though there’s no set timeline on when that will occur. The ads can be further spruced up with text overlays in a dozen of Meta’s most popular typeface options, while an automated image expansion tool — already used to adjust ads to fit different Meta surfaces like the TikTok lookalike Reels — will ensure the copy is aligned properly. 

Meta at the same time is expanding its AI-powered text generation to ad headlines in addition to primary text. The company said it factored in feedback from advertisers that wanted more diverse suggestions and ones that better reflected their brand values and product selling points. Outputs could be further improved as Meta transitions to its more advanced Llama 3 large-language models. 

The social media giant’s new AI bells and whistles are expected to be available globally by the end of 2024. As with many pitches around AI, the goal is to get rid of grunt work and open up more time for performance-driving activities, including creative versioning, that boost ad activity and, ultimately, Meta’s bottom line. Meta’s ad revenue was up 27% in Q1 while the average cost per ad increased 6%, a sign of growing demand. 

“We want to put time back into people’s hands who are actually focusing on strategic relationships ... as opposed to spending an inordinate amount of time transacting on our platform,” said Bowles.

Striking a balance

Most Meta advertisers have used AI in some shape or form for years, and the firm in 2023 released an AI Sandbox that leverages generative AI to let brands test out different backgrounds and text in their campaigns. Companies like Meta are now trying to gauge where more hands-off automation makes sense in advertising and what areas are best left under human supervision. 

Generative AI has been known to produce bizarre, off-putting images — Facebook is rife with uncanny valley material — and is increasingly subject to questions around copyright and the need for disclosure, such as watermarking that signals a picture was AI-generated. Meta executives were asked about both topics several times during the presentation. The Meta announcements also dropped the same day that OpenAI, a competitor in the tech arms race, announced a new Media Manager that gives artists a way to opt out of having their works used as training material for generative AI.   

Verticals like political and social issues advertisers will not be allowed to use Meta’s generative AI tools at first, per Hegeman. Meta in February outlined its plans to label AI-generated images, though its rules for advertisers are still being ironed out. 

“We’re in the process of working through some of the specifics in terms of exactly how that works in the context of ads,” said Hegeman. 

Meta will not put preference on AI-generated ads versus those made by humans, nor will it boost its own AI-generated ads over those created with rival software like Midjourney, Hegeman explained. The aim is to prioritize what’s performing best because that drives up the competitiveness of ad auctions. Asked about the cost of accessing these products, Hegeman assured that Meta tries to keep as many of its ad offerings free as possible. 

As generative AI becomes more sophisticated, and demands increased computing power, that could change. Meta leadership has previously indicated that charging people to use bigger AI models is one way to monetize the incredibly cost-intensive technology. At least for now, Meta is welcoming advertisers of all sizes into the fold.

“We really want to make sure that businesses of all sizes can option into this,” said Bowles. “This is really meant to democratize storytelling from an advertising lens.”

Apple unveiled a new generative artificial intelligence (AI) offering called Apple Intelligence during the keynote of its 2024 Worldwide Developers Conference (WWDC). The company also revealed a much-anticipated partnership with OpenAI.

Apple Intelligence is billed as a personal intelligence system for iPhone, iPad and Mac that joins generative models with personal context to drive relevancy. The offering, which promises to simplify everyday tasks and actions across a variety of apps, was introduced alongside a privacy-focused solution called Private Cloud Compute.

Apple is also integrating OpenAI’s ChatGPT into experiences within iOS 18, iPadOS 18 and macOS Sequoia, including within offerings like its new Writing Tools and Siri. The integration will be powered by OpenAI’s GPT-4o model and become available later in 2024.

Apple’s annual developer conference is known for its splashy displays of the latest and greatest the tech giant has to offer, including the newest innovations for its iOS software. Its latest event was no different, with presenters spending much of the event’s first-day keynote debuting Apple’s answer to the explosive rise of generative AI with the reveal of an OpenAI partnership and the unveiling of Apple Intelligence. The latter will be available this fall in beta on the latest versions of its operating systems, including iOS 18.

“Apple Intelligence will transform what users can do with our products — and what our products can do for our users,” said Apple chief executive Tim Cook in a statement. “Our unique approach combines generative AI with a user’s personal context to deliver truly helpful intelligence. And it can access that information in a completely private and secure way to help users do the things that matter most to them.” 

Apple is searching for its edge in the competitive AI landscape by promoting Apple Intelligence as a more individualized solution compared to other options, capable of understanding “personal context to deliver intelligence that is helpful and relevant.” For instance, in Apple’s Mail app, a new Priority Messages feature at the top of a user’s inbox will highlight more urgent emails, like a boarding pass. Within a user’s inbox they will also be able to see summaries of an email without needing to open a message, and a Smart Reply feature will suggest quick responses. 

Apple Intelligence will be integrated into several other Apple features. In the Notes and Phone apps, users will be able to record, transcribe and summarize audio. New Writing Tools within Apple’s latest operating systems will also leverage the tech. For instance, with a Rewrite function, Apple Intelligence will allow users to choose from different versions of what they have written and adjust the tone to suit the task at hand.

Apple is attempting to relieve any consumer privacy concerns around its AI — and maintain its walled-garden status — by designing it to feature on-device processing, with many of the models that power it running entirely on device. For more complex requests, the company introduced Private Cloud Compute, which draws on larger, server-based models that run on servers powered by Apple silicon in order to prevent data from being retained or exposed.

Notably, Apple will bring ChatGPT to experiences within iOS 18, iPadOS 18 and macOS Sequoia later this year as part of a highly anticipated partnership to allow consumers to utilize the tech without having to jump between tools. Within Siri, the virtual assistant can tap into ChatGPT’s expertise when deemed helpful, asking users for permission before any questions are passed along. ChatGPT will also be available in Apple’s Writing Tools to help users generate content for things they’re writing about. Through a Compose tool, users can access ChatGPT image tools to generate images in a swath of styles. 

Privacy protections are built in for Apple users who access ChatGPT: OpenAI won’t store requests, and IP addresses are obscured, per press details. 

Along with bringing ChatGPT to Siri, the keynote also detailed what is likely the largest overhaul to the virtual assistant to-date, describing the upgrades as its “new era.” Siri will be powered by Apple Intelligence to become more deeply integrated into the system experience, offering richer language-understanding capabilities that make it more natural and contextually relevant. Beyond Siri, the company announced a new Image Playground that leverages Apple Intelligence to allow users to create images within seconds and is accessible in apps like Messages along with its own dedicated app.

Among other announcements, Apple revealed that it will bring rich communication services (RCS) to Messages in iOS 18, a move the tech giant has long resisted, which will make for smoother conversations between iOS and Android devices. The company also announced that it will be bringing the Apple Vision Pro to new countries and a new feature for iOS 18, Messages via satellite, that enables messaging for the times when cellular and Wi-Fi connections aren’t available.

Dove has pledged to not use artificial intelligence (AI) to represent women in its advertising and communications as part of its long-running Real Beauty platform that is celebrating its 20th anniversary this year, according to a news release.     

An ad campaign, “The Code,” is set to a dramatic version of “Pure Imagination,” the song popularized by the film “Willy Wonka & the Chocolate Factory,” as women submit prompts to AI image generators to see what beauty looks like only to be greeted by unrealistic results. The video then shows Dove’s vision for what the technology could make, with a diverse array of women portrayed. 

The Unilever-owned brand has established a set of Real Beauty Prompt Guidelines that aim to help users of generative AI create more representative images. The brand enlisted the help of celebrity ambassadors including Reba McEntire, Drew Barrymore, Beanie Feldstein and Marsai Martin to raise awareness.

Dove describes AI as one of the “biggest threats” to portraying real beauty in its announcement, drawing a line between the emergent technology and the airbrushed imagery it has fought against since launching its signature messaging platform two decades ago. The brand’s new ad illustrates the potentially harmful nature of current AI tools as women sit in darkened rooms and see text prompts such as “a gorgeous woman” and “perfect skin” surface cartoonishly perfect results while a somber take on “Pure Imagination” swells.

All of the AI images that appear in the commercial were made with real generative AI tools, the company said. Brazil-based independent agency Soko developed the global marketing effort. The Real Beauty Prompt Guidelines and social and digital out-of-home aspects were handled by Media.Monks.

Dove is looking to differentiate from some beauty competitors by committing to not use AI-generated models in lieu of real women in its advertising moving forward. The Unilever mainstay believes AI could amplify the internet’s already negative impact on confidence, with nine in 10 women and girls surveyed by the brand saying they have been exposed to harmful beauty content online.

The press release and “The Code” ad also reference a forecast that claims 90% of online content will be AI-generated by 2025. A representative for Dove attributed the figure to author Nina Schick, who shared AI predictions as part of a CES panel discussion in 2023. Schick at the time said that 90% of online content may be at least partially created with AI by 2025. 

“At Dove, we seek a future in which women get to decide and declare what real beauty looks like — not algorithms. As we navigate the opportunities and challenges that come with new and emerging technology, we remain committed to protect, celebrate and champion Real Beauty,” said Alessandro Manfredi, CMO of Dove, in a statement. “Pledging to never use AI in our communications is just one step. We will not stop until beauty is a source of happiness, not anxiety, for every woman and girl.”

At the same time, Dove acknowledged that generative AI adoption is becoming difficult to ignore outright amid a surge in interest in software like ChatGPT. An internal study revealed that 24% of U.S. women and 41% of U.S. girls agreed that using AI to create different versions of themselves can be empowering. Dove hopes it can steer these enthusiasts in the right direction with the launch of the Real Beauty Prompt Guidelines that share inclusive language to get the most out of popular generative AI programs such as Midjourney and DALL-E.

While Dove identified some progress in broadening definitions of beauty since 2004, it emphasized there is still much work to be done. The marketer in partnership with Edelman’s DXI division assembled a report, titled “The Real State of Beauty,” that revealed over one in three women would trade a year of their life to realize their ideal appearance. Researchers spoke with over 33,000 respondents while extensive surveys were conducted across 20 countries.

As it prepares for a major relaunch, Toys R Us is embracing generative AI tools that have become the focus of the marketing and media industry. The company in June created what it claims is the first brand film to use OpenAI’s new text-to-video tool Sora, telling the story of how founder Charles Lazarus created the brand and its Geoffrey the Giraffe mascot, per a press release.

OpenAI kicked off the current AI frenzy with the launch of ChatGPT in November 2022 and is prepared to take another leap forward with the forthcoming release of its Sora tool. Sora, which is currently only available to alpha testers, can generate up to one-minute-long videos based on text prompts.

Toys R Us Studios relied on Native Foreign and Chief Creative Officer Nik Kleverov, who has alpha access to Sora, to create a one-minute photo-realistic video about the retailer’s origin. The video imagines how Charles Lazarus, the son of a bike-store owner, dreamed up Toys R Us and its iconic mascot. Toys R Us Studios President Kim Miller Olko served as executive producer and Kleverov acted as director on the project.

The full video, which is hosted on the brand’s website, includes a mention of the retailer’s pop-up stores in Macy’s locations. Parent company WHP Global in 2023 announced a partnership with Go Retail Group to open up to 24 flagship stores in the U.S. starting in 2024. Toys R Us filed for bankruptcy in 2017 before liquidating its stores in 2018, but now generates $2 billion in annual retail sales, according to WHP Global.

“The creative industry is experiencing a renaissance, much like Toys”R”Us,” Kleverov said in a statement. “Through Sora, we were able to tell this incredible story with remarkable speed and efficiency. Toys”R”Us is the perfect brand to embrace this AI-forward strategy, and we are thrilled to collaborate with their creative team to help lead the next wave of innovative storytelling.”

By using Sora, the brand and agency went from concept to final product in just a few weeks, condensing hundreds of iterative shots down to a few dozen. The video was created almost entirely with Sora but includes some corrective visual effects and an original score by Aaron Marsh of rock band Copeland.

Toys R Us is one of the many brands, including Coca-Cola and Starburst, that have used generative AI to create campaign materials. The technology is also key to growth plans for tech platforms like Google and Meta, as well as agency holding companies including Publicis, WPP and Havas.

WPP in June unveiled Production Studio, an application that uses generative artificial intelligence (AI) to create text, images and video for advertisers, according to a press release. The news was also announced at the 2024 Cannes Lions International Festival of Creativity. 

Co-developed with Hogarth, Production Studio provides direct access to WPP’s generative content engine supported by Nvidia and is implemented in the WPP Open operating system. The tool was built on Nvidia’s Omniverse development platform and relies on the chipmaker’s OpenUSD program for 3D workflows.

Generative AI was the talk of Cannes Lions as companies tried to position themselves on the leading edge of a trend that has shaken up marketing and media. WPP used the international advertising festival as an opportunity to promote its progress in implementing the emergent technology across its sprawling agency network, a strategy rivals are also pursuing. 

Production Studio takes advantage of a content engine recently developed with Nvidia and marks an important step in bringing generative AI tools to a larger number of WPP clients via the company’s Open operating system. Advertisers are hoping that generative AI can simplify what are often cost- and resource-intensive processes and allow for greater iterative capabilities with their creative assets, or, as WPP terms it, “exponentially more content.” For WPP, generative AI stands to be an important way to win business as growth flags and creative shops fall under sharper pressure. 

“Production Studio exemplifies WPP’s commitment to innovation and our ability to leverage cutting-edge technologies to drive growth for our clients through WPP Open,” said Stephan Pretorius, chief technology officer at WPP, in a statement. “By combining the power of AI with human creativity and expertise, we are redefining the boundaries of content creation and delivering value to our clients around the world.”

Production Studio assists in six key areas: Streamlining the production of 3D content models; producing high-quality, brand-relevant visuals; writing consistent, multilingual copy; compiling assets and ensuring they are tailored to the right channels; wedding assets with relevant data to improve targeting and personalization; and tracking performance to adapt asset performance quickly based on feedback. Nvidia’s OpenUSD helps generate hyper-realistic content at “an unprecedented volume,” according to the release. 

Production Studio has been piloted by brands including Ford and L’Oréal and is now available to all WPP clients. 

WPP said that Production Studio has guardrails and human oversight in place to ensure brand safety and legal and ethical compliance. Those are some of the top concerns surrounding generative AI, which is trained by scraping existing media. 

Production Studio is just one piece of WPP’s commitment to spend 250 million pounds, or about $318 million, annually on generative AI over the next several years. At Cannes Lions, the group also revealed a partnership with IBM around using generative AI to enhance its business-to-business marketing solutions. In late May, WPP brought the latest AI models of startup Anthropic into its operating system. Open previously integrated Google’s Gemini AI, OpenAI’s ChatGPT and DALL-E and a variety of other AI software. 

WPP is betting heavily on generative AI amid a larger turnaround push. The U.K.-based firm experienced revenue declines in Q1 2024, intensified by the loss of some major accounts and broad tech spending pullbacks, and expects full-year growth to be flat or at 1%.