4A's Noteworthy News - 12th Edition
THE BIG STORY: Walking the GenAI Tightrope - Balancing Transparency and Trust in Advertising
Recent articles reveal a critical balancing act underway in the advertising industry: transparency in AI usage is crucial to maintaining trust, but disclosing AI involvement risks alienating consumers.
A study by Washington State University found that using the term "artificial intelligence" in product descriptions notably reduces purchase intentions. This indicates that despite AI’s potential benefits, consumers are wary of its integration into products and services.
Furthermore, a candid discussion with an agency executive via Digiday highlighted how generative AI - often heralded as a revolutionary tool for efficiency, productivity and innovation - is causing client apprehension. The introduction of “Made with AI” labels on platforms like Facebook and Instagram has led to concerns about the authenticity and trustworthiness of content. Clients fear that these labels might suggest their products are less genuine or artificially enhanced, undermining consumer confidence.
The executive explained that even minor AI usage, such as image enhancements in Photoshop, triggers these labels, leading to a credibility crisis. This situation forces agencies to rethink their use of AI-powered tools, balancing benefits with the risk of eroding client and consumer trust.
It is yet another piece of the responsible AI puzzle, and a bit of a conundrum that will not be easily solved. In addition to platform rules already in place, regulation is coming that may legally require disclosure for certain AI use cases. The 4A’s is actively examining the issue in partnership with our member agencies. Meantime, agencies must carefully navigate the AI landscape, ensuring their use of technology does not alienate the very audience they aim to engage.
Recently released chapter 5 of our Generative AI Blueprint (a members-only resource) provides helpful frameworks and guidance for AI Governance and Risk Management.
OTHER INTERESTING AI NEWS FOR AGENCIES:
Related to the big story, the U.S. Copyright Office tells Congress: there is an 'Urgent need' to outlaw AI-powered impersonation.
Gartner Predicts Ongoing Challenges with AI projects, forecasting that 30% of generative AI projects will be abandoned after proof-of-concept by 2025 due to poor data quality, inadequate risk controls, escalating costs, and unclear business value.
Dentsu’s 2024 CMO report finds shifting perceptions on human + machine creativity.
Ad Age’s AMP Platform invites agency execs to share how AI is changing the nature of their work.
Perplexity formally launched the recently announced Perplexity Publishers’ Program, designed to share ad revenue with publishers and creators.
After demoing a polarizing new advanced voice capability in May, OpenAI delayed the rollout, citing safety and security concerns. It is now beginning to roll out to select users in an alpha program.
As LLMs run out of publicly available data to train on, synthetic data generated by other AI models has been positioned as a potential solution. Results are mixed, however, and a new study published in Nature does not bode well.
Google introduced updates to its small, open-ish model Gemma 2 - as the march towards smaller on-device models continues.
RECENT AI CAMPAIGNS AND AGENCY USE CASES:
An IPG-owned production studio leveraged AI in developing its new brand.
New Nvidia microservices are expected to automate production steps for WPP creative teams.
NBCUniversal introduces “OLI,” an AI-powered logic system that helps viewers find when, where and how to watch their favorite live events and programming throughout NBCU’s coverage of the 2024 Paris Olympics.
NOTE: This article first appeared on LinkedIn as part of a regular series I’m writing for the 4A’s.
