Why advertising experts say new AI protocol won't fix programmatic's biggest problem

Two advertising technology experts debated the future of automated ad buying on LinkedIn during December 2025, revealing why a new industry protocol might fail before it starts. The discussion centered on a fundamental question: should artificial intelligence agents buy ads directly from publishers, or should they continue using the programmatic advertising system that currently verifies billions of dollars in transactions each day?

Christopher F., who builds advertising platforms at Attention Arc, and Dr. Augustine Fou, a fraud investigator who has studied digital advertising for 30 years, both expressed skepticism about the Ad Context Protocol. The protocol, known as AdCP, launched on October 15, 2025, with promises that AI agents could buy advertising directly across platforms without the current programmatic infrastructure.

The debate highlights a larger tension in digital advertising. Companies spent approximately $700 billion globally on programmatic advertising in 2024. This system uses supply-side platforms as gatekeepers between publishers who sell ad space and advertisers who want to buy it. SSPs verify transactions, filter fraud, and manage auctions that happen in 100 milliseconds as web pages load.

AdCP wants to bypass much of this system. According to its creators, AI agents could communicate directly with publishers to discover available ad space, compare prices, and place ads automatically. Six companies launched the protocol: Scope3, Yahoo, PubMatic, Swivel, Triton Digital, and Optable.

How programmatic advertising works today

When someone visits a website, programmatic advertising technology runs an instant auction for the ad space. Publishers use supply-side platforms to connect their available ad inventory to potential buyers. Advertisers use demand-side platforms to bid on this inventory based on who they want to reach.

SSPs serve several critical functions in this process. They manage which advertisers can bid on publisher inventory. They filter out fraudulent traffic before it reaches advertisers. They provide transparency about where ads appear and how much money flows through the system. They prevent the same advertiser from bidding against itself when the same ad space appears through multiple channels.

"SSPs stand at unique positions in the advertising supply chain where proximity to publishers and direct inventory access enables more efficient programmatic transactions," according to analysis from Equativ published on October 24, 2024. SSPs can pre-qualify inventory and consolidate targeting information, making the buying process simpler for advertisers.

This system emerged because direct connections between thousands of advertisers and thousands of publishers proved unworkable. The technology infrastructure required to handle billions of transactions per day, verify quality, prevent fraud, and settle payments exceeded what most individual companies could build.

Why AdCP faces business obstacles

Christopher F. posted his analysis on LinkedIn explaining why major advertising platforms will resist adopting AdCP. "That statement is not wrong. In fact, it is refreshingly honest," he wrote, responding to commentary about the protocol's benefits.

The commentary had noted that AdCP could eliminate internal APIs that represent custom development, engineering investment, and ongoing maintenance costs. These internal systems create vendor lock-in, making it difficult for customers to switch platforms.

Christopher F. acknowledged this problem exists. Adopting any new protocol costs money, especially for large companies processing millions of transactions per second. Companies only make such investments when they clearly solve problems or demonstrate revenue growth potential.

"Vendor lock-in, while frustrating for customers seeking change, remains undeniably beneficial for business operations," Christopher F. wrote. "The harder it becomes to switch platforms, the less likely customers are to do so."

He identified three major advertising platforms that dominate the open web market in the United States. These platforms have little incentive to adopt a protocol that lowers switching costs, requires new technology, and introduces risks where few currently exist.

"Does anyone genuinely believe Google will adopt a protocol that makes it easier for dependent trading desks and customers to move spend to a competitor?" Christopher F. asked in his December 2025 post.

His assessment focused on who makes adoption decisions. Marketing executives might like the idea of unified protocols that simplify workflows. Technology executives who must build and maintain these systems face different calculations. "Protocol adoption depends on convincing chief technology officers rather than chief marketing officers," he noted.

Christopher F. discussed these dynamics during a Digiday talk available on their podcast feed. He argued that until protocol proponents address actual business incentives, "hype is likely to outpace actual adoption for considerable time."

Why AdCP faces fraud obstacles

Dr. Augustine Fou responded with concerns drawn from his fraud investigation work. "The protocol itself contains merit in theory, similar to many standards that preceded it," he wrote on LinkedIn in December 2025. "But what it doesn't take into account, like many standards before it, are attack scenarios where fraudsters exploit it to commit MORE fraud."

Dr. Fou cited ads.txt as an example. The IAB Tech Lab created ads.txt in 2017 to combat unauthorized inventory sales. Publishers place a text file on their website listing companies authorized to sell their ad space. Advertisers check this file to verify legitimate sellers.

"The standard represented sound thinking in theory, but from its first day of release, fraudsters used it to commit more fraud and provide better cover for their operations," Dr. Fou wrote. Criminals created fake publisher websites with ads.txt files declaring fraudulent sellers as authorized. They also compromised legitimate publisher websites to add unauthorized sellers to existing ads.txt files.

The problem stems from how ads.txt works. The system relies on publishers declaring accurate information. It provides no independent verification that declarations are truthful. Fraudsters exploited this trust assumption.

A February 2024 Pixalate report found that 25% of programmatic traffic with SupplyChain Object data failed validation checks. SupplyChain Object, known as sellers.json, works similarly to ads.txt. It enables buyers to verify seller legitimacy by checking declarations in bid requests against what supply-side platforms report.

Traffic that failed these validation checks had 64% higher invalid traffic rates compared to traffic that passed validation, according to the Pixalate research examining Q4 2023 data. "Our latest SCO report highlights the importance of transparency in the supply chain and the need for vigilance when depending on SCO," stated Amit Shetty, vice president of product at Pixalate.

AdCP faces the same trust problem. "AdCP relies on parties to declare things correctly and honestly," Dr. Fou stated. "As we know from the last 15 yrs, that's not going to happen."

What SSPs do that AdCP cannot

Supply-side platforms evolved beyond simple inventory connections to provide verification services that protect the advertising ecosystem. These functions become more critical as fraudsters develop sophisticated attack methods.

Industry anti-fraud efforts saved advertisers $10.8 billion in 2023 according to research from the Trustworthy Accountability Group, 4A's, Association of National Advertisers, and Interactive Advertising Bureau. The study found that without anti-fraud programs, invalid traffic would have cost the US advertising industry $11.78 billion based on a 9.96% invalid traffic rate.

TAG's Certified Against Fraud program maintains invalid traffic rates below 1% across certified channels globally. This represents a 92% reduction compared to what losses would be without these protection systems.

SSPs participate in this fraud prevention ecosystem through multiple mechanisms. They implement pre-bid filtration that blocks fraudulent traffic before auctions occur. They maintain post-serve verification that identifies fraud after ads deliver. They participate in threat intelligence sharing that helps the industry respond to emerging attack patterns.

European anti-fraud programmes prevented €3.45 billion in losses during 2023, representing 69% reduction in fraud-related costs, according to a June 2025 TAG study. The research examined European video and display advertising channels worth €48.3 billion.

Channels where companies achieved TAG certification accounted for 76% of this spending, experiencing fraud losses of just €115 million at a 0.3% invalid traffic rate. The remaining 24% of spending flowed through uncertified channels that sustained €1.19 billion in fraud losses, more than ten times the relative level compared to protected channels.

The data comes from four Media Rating Council-accredited invalid traffic vendors: DoubleVerify, Integral Ad Science, Moat by Oracle, and Pixalate. These companies provided measurements from 2023 and early 2024 establishing the baseline unfiltered fraud rate of 10.36% in unprotected channels.

AdCP contains security considerations including authorization verification and fraud prevention mechanisms, according to technical documentation. However, these rely primarily on participants declaring information accurately, similar to the trust model that fraudsters exploited with ads.txt.

How the current system verifies transactions

Real-time bidding auctions occur within approximately 100 milliseconds as web pages load. Multiple advertisers submit bids simultaneously based on available audience and contextual information. The highest bidder wins the right to display their advertisement.

SSPs connect publishers to these auctions, providing tools for inventory management, yield optimization, and revenue reporting across digital properties. They aggregate inventory from multiple publishers, enabling advertisers to access diverse ad placements through unified platforms.

The platforms also prevent request duplication, where the same advertising opportunity appears multiple times through different supply paths. Transaction IDs enable demand-side platforms to recognize when different bid requests represent the same ad space, preventing advertisers from bidding against themselves.

Prebid.org removed this cross-exchange tracking capability on August 27, 2025, following publisher demands for supply path control. The change eliminated transparency tools that helped advertisers detect duplicate bids, according to industry analysis. The IAB Technology Laboratory declared the implementation violated OpenRTB specifications.

This incident demonstrates ongoing tensions between publisher yield optimization and advertiser transparency. Publishers want to maximize revenue by selling the same inventory through multiple channels. Advertisers want to avoid paying multiple times for the same ad space.

SSPs navigate these tensions through supply path optimization strategies. Advertisers evaluate supply-side platforms, ad exchanges, and publisher connections to eliminate unnecessary technology fees and latency. According to analysis, match rate losses between platforms typically range from 40% to 70%, making direct supply path connections valuable for maintaining data fidelity.

What experts say about protocol adoption

Industry veteran Ari Paparo questioned AdCP's viability on November 3, 2025, while expressing support for creative automation specifications. He raised concerns about whether AI agents can overcome fundamental business problems including pricing transparency resistance, scale economics, and data fragmentation that prevented programmatic direct success historically.

"So again, the use of agents ends up being more appropriate for reinforcing the role of intermediaries, rather than enabling a breakthrough to empower buyers and sellers," Paparo wrote in his analysis. Long-tail publishers lack technical sophistication and data scale necessary to differentiate themselves to advertisers without aggregation.

The protocol may still deliver value, but likely to different constituencies than intended. "It's just that the value will likely accrue to very large publishers and cross-publisher ad networks, in much the same way it has with programmatic more generally," Paparo stated.

David Kohl warned that AdCP represents "the tail wagging the dog" in an October 15, 2025, LinkedIn post. He cautioned that rushing to build agentic advertising infrastructure without clear goals risks repeating mistakes that created today's dysfunctional programmatic supply chain.

Kohl argued that the programmatic supply chain evolved without foresight or a master plan, creating an ecosystem where intermediaries prioritized profits over advertiser and publisher needs. He called for shared goals, structured innovation approaches, and measurable success metrics before committing to specific technical implementations.

The criticism matters because transparency issues resulted in only 36% of post-transaction programmatic budgets reaching valid, viewable, measurable impressions according to 2023 research.

Why previous protocols failed

The advertising industry has witnessed multiple standardization efforts face adoption obstacles despite technical merit. OpenRTB 3.0 provides an instructive example.

OpenRTB became the foundation of programmatic advertising, establishing how real-time bidding communications work between buyers and sellers. The 3.0 version aimed to improve on earlier specifications with enhanced features and better data models.

However, OpenRTB 3.0 faced adoption challenges because implementation required complete recoding of existing systems. Companies evaluated the engineering investment required and many declined to adopt the new version, continuing to use earlier specifications instead.

This pattern reveals the coordination problem that protocol adoption faces. Protocols only work when enough market participants implement them. Engineering teams must justify the costs of building support for new standards. Business leaders must see clear return on investment.

Simpler protocols gain adoption more easily. Ads.txt succeeded at widespread implementation because publishers could add it by placing a text file on their domain. The technical barrier remained low enough that even small publishers could participate.

However, simplicity created the vulnerability that Dr. Fou identified. Systems relying on self-declaration without verification become targets for fraud. Criminals create false declarations that appear legitimate until sophisticated detection systems identify the deception.

What Christopher F. builds to stop fraud

Christopher F. brings direct experience with these challenges through his work building advertising platforms and fraud detection systems. As Director of Product Development at Attention Arc and Head of Media Product at Cheil Worldwide, he architected A+, described as an enterprise-grade agency trading platform.

The system uses Golang programming language running on Ubuntu Linux with dedicated microservices for data collection and anomaly detection. These components work together to identify fraudulent traffic patterns before they impact advertising campaigns.

Prior to his current role, Christopher F. founded Spartan Ad in April 2021, building an enterprise-level ad fraud prevention platform. The system also uses Golang with the Gin-Gonic framework, deployed on scalable Linux-based infrastructure.

His patent portfolio includes "Fraud Prevention in Programmatic Advertising," filed September 10, 2019, patent number 11,521,231. The patent describes machine learning-based systems and methods for preventing fraud in programmatic advertising environments.

Between September 2015 and January 2021, he served as chief product officer at a company where he helped grow the organization from zero revenue to $150 million valuation in four years. He designed and developed the real-time bidder used in digital advertising for the company.

Building these systems requires understanding how fraudsters attack programmatic infrastructure. Invalid traffic takes many forms, from simple bots that click on ads to sophisticated operations that spoof premium publisher inventory.

How Dr. Fou investigates fraud patterns

Dr. Augustine Fou has documented advertising fraud patterns since founding FouAnalytics, marketed as an alternative to mainstream analytics for verifying digital advertising campaigns. His background includes a PhD from Massachusetts Institute of Technology in Materials Science and Engineering, completed at age 23.

He previously worked at McKinsey & Company before serving as Group Chief Digital Officer at Omnicom's Healthcare Consultancy Group, a $100 million organization with eight agencies serving pharmaceutical, medical device, and healthcare clients.

His fraud investigation work involves assisting government and regulatory bodies while helping commercial clients strengthen cybersecurity and mitigate threats. According to his LinkedIn profile, he has been on the front lines of digital marketing for nearly three decades.

FouAnalytics operates as a free platform for small businesses to verify digital ad campaigns and clicks from various paid online sources. The system provides detailed analysis of traffic sources, enabling advertisers to identify fraudulent patterns that inflate costs without delivering real engagement.

Dr. Fou taught digital strategy and integrated marketing at Rutgers University and NYU's School of Continuing and Professional Studies. He completed his MIT PhD thesis on "Design, Fabrication, and Characterization of Complex Multilayer Thin Films of Conducting and Electroluminescent Organic Polymers through Molecular Self-Assembly."

His technical background informs his fraud analysis approach. He understands how complex systems work and how adversaries exploit weaknesses in their design assumptions. The ads.txt example he cited illustrates this pattern-finding capability.

Why major platforms avoided AdCP

The six companies that launched Ad Context Protocol represent various segments of advertising technology. PubMatic operates as a supply-side platform. Scope3 focuses on carbon emissions measurement and agentic advertising. Yahoo brings media properties and advertising infrastructure. Swivel, Triton Digital, and Optable provide targeting, audio advertising, and data collaboration capabilities.

Notably absent from this list: Google, Amazon, The Trade Desk, and Microsoft. These companies operate the dominant demand-side platforms that process the majority of programmatic advertising budgets.

Lindsay Rowntree, COO at ExchangeWire, noted during a podcast discussion that major platforms including Google, The Trade Desk, and Amazon DSP have not signed up for recent agentic protocols. This potentially creates more fragmentation through new walled gardens rather than solving existing coordination problems.

Google operates the largest advertising platform globally, processing billions of transactions daily. The company has minimal business incentive to support protocols that make customer migration easier. Its advertising technology stack integrates tightly across Google Ads, Display & Video 360, Google Ad Manager, and YouTube.

Amazon DSP expanded rapidly throughout 2024 and 2025, adding supply-side platform partnerships with Disney in June 2025, Spotify in October 2025, and Netflix in Q4 2025. Microsoft Monetize joined as a preferred partner in October 2025, creating new supply path connections before Microsoft Invest shutdown on February 28, 2026.

The Trade Desk launched OpenAds platform on October 2, 2025, after Prebid.org removed transaction ID functionality. The platform aims to maintain transparency tools that advertisers lost through the Prebid changes. This demonstrates The Trade Desk's strategy of building proprietary solutions rather than adopting industry-wide protocols it does not control.

What happens next

Multiple agentic AI protocols emerged during fall 2025. LiveRamp donated the Universal Context Protocol to IAB Tech Lab on November 3, 2025. IAB Tech Lab introduced its Agentic RTB Framework v1.0 for public comment on November 13, 2025.

Each new protocol generates industry discussion about whether advertising needs additional standards before addressing fundamental problems. The proliferation itself creates coordination challenges as companies evaluate which frameworks to support with limited engineering resources.

Microsoft announced on May 14, 2025, that it would discontinue Microsoft Invest effective February 28, 2026. The company cited incompatibility between traditional demand-side platform models and their vision for "conversational, personalized, and agentic" advertising futures, according to Microsoft Advertising Corporate Vice President Kya Sainsbury-Carter.

Amazon merged its DSP and Ads Console into unified Campaign Manager at the company's annual unBoxed conference on November 10, 2025. An agentic mode called Full-Funnel Campaigns sets up and adjusts multi-format campaigns across Sponsored Products, Sponsored Brands, display, and streaming television from a single prompt.

These developments suggest major platforms are building agentic capabilities within their own ecosystems rather than adopting cross-platform protocols. This approach maintains the vendor lock-in dynamics that Christopher F. identified as obstacles to AdCP adoption.

The LinkedIn discussion between Christopher F. and Dr. Augustine Fou crystallized tensions that will determine how advertising automation develops. Protocols need widespread adoption to function. Adoption requires convincing technology executives that business benefits justify engineering investment and risk.

Simultaneously, protocols must account for fraud exploitation scenarios. Systems relying on self-declaration without independent verification become targets for criminals. The balance between simplicity that enables adoption and complexity that enables verification remains unsolved.

Supply-side platforms occupy critical positions in this ecosystem. They verify transactions, filter fraud, optimize yield for publishers, and provide transparency tools for advertisers. Whether AI agents can replicate these functions without concentrated intermediaries remains uncertain.

What seems clear from the December 2025 debate: AdCP faces significant obstacles from both business incentives and security requirements that its design does not adequately address.

Timeline

• Mid-2010s: Header bidding emerged, making leading SSPs interchangeable for basic inventory access
• 2017: IAB Tech Lab created ads.txt to combat unauthorized inventory sales; fraudsters began exploiting it immediately according to Dr. Augustine Fou
• September 10, 2019: Christopher F. filed patent for fraud prevention in programmatic advertising
• April 2021: Christopher F. founded Spartan Ad, enterprise-level ad fraud prevention platform
• December 2022: Christopher F. became Head of Media Product at Cheil Worldwide
• Q4 2023: 25% of programmatic traffic with SupplyChain Object failed validation checks, with 64% higher invalid traffic rates
• 2023: Industry anti-fraud efforts saved advertisers $10.8 billion through 92% reduction in invalid traffic losses
• 2023: European anti-fraud programmes prevented €3.45 billion in losses, representing 69% reduction in fraud costs
• May 14, 2025: Microsoft announced discontinuation of Microsoft Invest effective February 28, 2026
• July 2025: Christopher F. discussed AdCP adoption challenges during Digiday talk
• August 27, 2025: Prebid.org removed cross-exchange transaction ID functionality
• September 2025: Index Exchange became first SSP to integrate Gracenote contextual intelligence
• September 2025: Best Buy selected Magnite as exclusive SSP to expand programmatic access
• October 2, 2025: The Trade Desk launched OpenAds platform to maintain transparency after transaction ID changes
• October 7, 2025: Amazon DSP integrated Microsoft Monetize SSP as preferred partner
• October 15, 2025: Six companies launched Ad Context Protocol for advertising automation
• October 15, 2025: David Kohl posted LinkedIn criticism warning AdCP is "the tail wagging the dog"
• November 3, 2025: LiveRamp donated Universal Context Protocol to IAB Tech Lab
• November 3, 2025: Ari Paparo published analysis questioning AdCP media buying protocol viability
• November 10, 2025: Amazon merged DSP and Ads Console into unified Campaign Manager with AI agents
• November 13, 2025: IAB Tech Lab introduced Agentic RTB Framework v1.0 for public comment
• December 2025: Christopher F. and Dr. Augustine Fou debated AdCP viability on LinkedIn

Summary

Who: Christopher F., Director of Product Development at Attention Arc and Head of Media Product at Cheil Worldwide, and Dr. Augustine Fou, advertising fraud investigator with FouAnalytics, debated the Ad Context Protocol on LinkedIn. Six companies launched AdCP: Scope3, Yahoo, PubMatic, Swivel, Triton Digital, and Optable. Industry experts including Ari Paparo and David Kohl also criticized the protocol.

What: The debate centered on whether the Ad Context Protocol can replace current programmatic advertising infrastructure that uses supply-side platforms to verify transactions and filter fraud. AdCP aims to enable AI agents to buy advertising directly from publishers without intermediaries. Critics argue major platforms lack business incentive to adopt it, and the protocol relies on self-declaration that fraudsters will exploit, similar to how ads.txt was subverted from its first day.

When: The LinkedIn discussion occurred in December 2025, following AdCP's October 15, 2025, launch. The debate emerged as multiple agentic AI protocols launched during fall 2025, including Universal Context Protocol on November 3 and Agentic RTB Framework on November 13. Microsoft plans to shut down Microsoft Invest on February 28, 2026, citing incompatibility with agentic advertising visions.

Where: The discussion occurred on LinkedIn but addressed the global programmatic advertising ecosystem worth approximately $700 billion in 2024. Supply-side platforms process billions of transactions daily across display, video, mobile, and connected television environments. The debate affects advertisers, publishers, and technology platforms operating across the open web market.

Why: The debate matters because it reveals fundamental obstacles facing advertising automation. AdCP proponents claim protocols reduce custom development costs and vendor lock-in. Critics argue dominant platforms benefit from customer switching costs and will resist adoption. Meanwhile, fraud prevention systems maintained by SSPs saved advertisers $10.8 billion in 2023 through invalid traffic filtration. Protocols relying on self-declaration without verification create new fraud attack vectors, as demonstrated by ads.txt exploitation. Only 36% of programmatic budgets reach valid, viewable, measurable impressions according to 2023 research, suggesting automation must address transparency before adding complexity.