The Programmatic Digest podcast

207. The AI Middle Layer: The Next Brand Safety Battleground

0:00
44:35
Rewind 15 seconds
Fast Forward 15 seconds

In this podcast episode, Dr. Cecilia Dones and Iesha White discuss the complex evolution of artificial intelligence within the marketing and advertising industries. 

They explore the shift of AI becoming a "middle layer" between brands and consumers, highlighting how automated systems can influence purchasing decisions using unverified information. 

These outstanding ladies distinguish between closed and open AI systems, warning that proprietary models often lack transparency and can lead to brand safety issues like hallucinations or data poisoning. 

They emphasize that while AI offers efficiency, brands must prioritize human judgment, organic engagement, and authentic connections to maintain their reputation. Ultimately, the conversation serves as a call for marketers to implement stronger governance and return to fundamental marketing principles to navigate an increasingly automated landscape.

Here are four topics we focused on:

  • The AI "Middle Layer": AI is increasingly acting as an unapproved intermediary between brands and consumers. AI systems now generate answers, ad creative, and buying decisions using unverified information that brands did not write or approve.

  • The Trap of Closed Systems: Major ad networks heavily push automated, "closed" AI tools (like Google's PMAX and Meta's Advantage+) to recoup their massive infrastructure investments. This leaves brands with an "illusion of choice" and little-to-no technical ability to trace, audit, or correct costly AI hallucinations.

  • SEO Poisoning & Sabotage: Closed systems are highly vulnerable to model manipulation. A Cornell study showed it takes as few as 13 words on Reddit to completely alter Google's AI Overviews. This enables black-hat corporate espionage where bad actors can programmatically poison what LLMs say about a competitor.

  • Defense & Adaptation: Brands must collectively lobby agencies and industry groups to demand LLM transparency. Meanwhile, marketers should prioritize core organic SEO and human-driven curation, while consumers are increasingly pushing back against digital clutter by seeking offline, real-world connections ("touch grass")  

Check out the interview on YouTube! 

 

 

🧠 About Us:

We are Helene Parker Consulting, where programmatic hands on keys buying meet Revenue Growth. We're expanding our work with Independent agencies, media and ad-tech companies through revenue growth, go-to-market support, pipeline development, and sales enablement. We also offer practical programmatic workshops and customized training for teams that want to strengthen their skills, with a trader/buyer lens.

Are you or anyone in your network looking for support? Email us [email protected]

Past workshops, paid gems, and every new recording + resource added as we go. This library isn't a one-time thing, it's the full vault plus all future trainings, recordings, and guides 

 Access it here: https://www.heleneparker.com/library

 

 

⏱ Timestamps:

00:00 Introduction and welcome

01:06 Guest introductions: Dr. Cecilia Dones and Aisha White

02:15 Why AI is becoming the middle layer between brands and consumers

04:27 Current AI developments and the importance of watching patterns

07:01 Open versus closed AI systems explained

10:13 AI errors and loss of control in digital advertising

13:00 Using AI automation without surrendering control

16:02 Why platforms are pushing advertisers toward AI

17:50 AI agents and advertising to machines

19:18 What an AI agent is and how it works

21:02 AI agents, identity, and consumer privacy

22:56 Will marketers need separate strategies for humans and AI agents?

23:48 Platforms restricting outside AI agents

25:35 Walled gardens and machine-to-machine advertising

28:01 AI hallucinations, feedback loops, and brand reputation

29:05 Why SEO, organic content, and marketing fundamentals still matter

32:16 Connecting digital marketing with real-life experiences

34:39 Expanding brand safety to include what AI says about your brand

36:08 How manipulated content can poison AI results

39:18 Practical AI protection advice for brands and agencies

41:29 Advice for consumers navigating AI-generated information

43:10 Final thoughts and guest contact information

 

👤 Meet Our Guests:

Dr. Cecilia Dones - Founder & Chief Data Officer, Three Standard Deviations

Dr. Cecilia Dones is an educator and technology leader who runs Three Standard Deviations, where she focuses on helping organizations navigate responsible technology and AI implementations. She also writes the Substack Authentic Interactions, where she shares her musings regarding technology and the lived experience.

Subscribe to her Substack: https://substack.com/@ceciliadones

Website: https://www.3standarddeviations.com/ 

📎 Connect on LinkedIn: linkedin.com/in/ceciliadones 

 

Iesha White - Director of Intelligence, Check My Ads Institute

Iesha White is a programmatic advertising veteran with over a decade of experience in the industry. She has worked in both the United States and the United Kingdom, spanning roles on the ad-tech side, brand side, and agency side twice each. In her current role at Check My Ads Institute, she oversees all research and investigations. 

📎 Connect on LinkedIn: ​​https://www.linkedin.com/in/iesha-white-marketing/  

 

 

💻 Meet The Team:

Hélène Parker – Chief Programmatic Coach

🌐 heleneparker.com
🔗 LinkedIn: https://www.linkedin.com/in/helene-parker/

Manuela Cortes – Co-Host, Programmatic Digest en Español

🔗 LinkedIn: https://www.linkedin.com/in/manuela-cortes-/

 

 

🎓 Learn Programmatic:

For Individual Buyers: https://heleneparker.com/library

Subscribe to the Newsletter: https://heleneparker.com/newsletter

Be the first to find out about free programmatic training workshops: https://www.heleneparker.com/waitlist/

 

 

📲 Follow Us:

🔗 LinkedIn: https://www.linkedin.com/company/programmatic-digest-podcast

▶️ YouTube: https://www.youtube.com/@programmaticdigest

 

More episodes from "The Programmatic Digest"