Ad Context Protocol divides advertising industry on agentic AI standards

A coalition of advertising technology companies launched the Ad Context Protocol earlier this month, betting that a new open-source technical standard could enable artificial intelligence agents to communicate across platforms and execute advertising tasks autonomously. The October 15 announcement drew immediate scrutiny from industry observers questioning whether the sector needs another protocol when existing standards remain underutilized.

The protocol emerged from a group including Scope3, Yahoo, PubMatic, Swivel, Triton, and Optable. According to the companies involved, AdCP provides a unified interface that allows AI agents to discover inventory, compare pricing, and activate campaigns across different advertising platforms without requiring custom integration work for each system.

Brian O'Kelley, CEO of Scope3 and a founding member of AdCP, addressed concerns about agent trustworthiness in a blog post. "AdCP requires publishers to register sales agents publicly," O'Kelley wrote. The protocol mandates that buyers verify publisher domains to confirm what the AI agents are authorized to sell.

Defenders of the protocol argue the shift to agentic advertising represents an inevitable transformation. Ben Kahan, senior director of programmatic at the media agency Brainlabs, told The Current: "The industry goes through eras, and now we seem to be entering an agentic era. But everything is siloed or fragmented. There hasn't yet been a push for standardization across any of it."

Existing protocols complicate adoption

The launch arrives amid existing artificial intelligence standards, including Agent2Agent and Model Context Protocol. Anthony Katsur, CEO of IAB Tech Lab, questioned the timing. "What we don't need is another industry trade group," Katsur told The Current. He noted that IAB Tech Lab operates open-source structures where anyone can work, including Ads.Cert. "We already have solved some of the problems that this AdCP initiative is trying to," Katsur said.

The IAB Tech Lab maintains that it is developing new standards supporting agentic frameworks, with details expected in the coming months. Katsur acknowledged substance exists in agentic capabilities beyond typical industry trends. "We are a shiny-penny industry, and agentic is a shiny penny — but I don't think it's blockchain," he said. "There is something of substance there."

The protocol's structure covers three distinct phases. During discovery, users describe target audiences in natural language rather than navigating platform-specific targeting interfaces. The comparison phase presents pricing, reach, and targeting capabilities in consistent data structures. Activation launches campaigns across selected platforms through single commands while maintaining platform-specific optimizations.

According to technical documentation, AdCP provides nine core tasks covering the complete advertising lifecycle. The get_products function discovers advertising inventory using natural language campaign briefs. The create_media_buy function launches campaigns with budget allocation, timing parameters, and promoted offering specifications.

Transparency concerns persist

Augustine Fou, a fraud researcher and marketing consultant, cautioned that agentic AI does not eliminate bad actors in the supply chain. "More automation means less transparency," Fou told The Current. Agents can still act on behalf of people with bad incentives, he noted.

The protocol addresses long-standing fragmentation issues. Each advertising platform currently maintains proprietary application programming interfaces with distinct workflow requirements, documentation standards, and reporting formats. This forces media buyers and agencies to develop separate integration capabilities for Google Ad Manager, connected television platforms, and digital out-of-home systems.

Agentic AI became prominent at the ANA Masters of Marketing conference in Orlando, Florida, last week. Shelly Palmer, president and CEO of The Palmer Group, warned marketers to "beware" of AI agents before implementing them into workflows. "Agents are intelligence decoupled from consciousness," Palmer said during a keynote. "They are literally synthetic employees."

Skepticism about commercial viability

Joe Root, co-founder of Permutive, expressed concerns that AdCP could distract from foundational performance issues. "If the product doesn't compete on incremental outcomes, it doesn't matter," Root told The Current. "AdCP runs the risk of publishers jumping to create agents when they haven't fixed the foundational problem."

Root suggested the protocol could distract the industry from matching performance with walled gardens like Google and Meta. John Hoctor, CEO and co-founder of Newton Research and an AdCP founding member, disagreed with this assessment. "Walled gardens haven't been shy about the fact that they're planning to use agents to help their clients plan, buy, measure and optimize campaigns," Hoctor said.

According to Root's conversations with publishers, no companies are forecasting transacting a single dollar against AdCP at present. He characterized the protocol as primarily allowing the industry "to congregate and get the ball rolling" at this stage.

Hoctor acknowledged that agents decisioning directly into the media buying process is "complicated." He suggested agents could perform other tasks like analytics and measurement in the near term while more complex implementations develop.

The advertising industry faces increasing pressure toward automation as campaign complexity grows. LiveRamp introduced agentic orchestration capabilities on October 1, enabling autonomous AI agents to access identity resolution, segmentation and measurement tools. Adobe launched its Experience Platform Agent Orchestrator on September 10 for managing agents across Adobe and third-party ecosystems.

Investment momentum builds despite doubts

McKinsey data indicates substantial investment momentum in agentic artificial intelligence. The consulting firm reported that $1.1 billion in equity investment flowed into agentic AI during 2024. Job postings related to the technology increased 985 percent from 2023 to 2024, according to McKinsey's Technology Trends Outlook 2025.

The timing coincides with significant market disruptions affecting transparency in programmatic advertising. Microsoft announced on May 14 that it would discontinue Microsoft Invest (formerly Xandr) effective February 28, 2026. According to Microsoft Advertising Corporate Vice President Kya Sainsbury-Carter, the company cited incompatibility between traditional demand-side platform models and their vision for "conversational, personalized, and agentic" advertising futures.

Industry analysis suggests autonomous AI systems could automate campaign setup, targeting and optimization functions currently handled by demand-side platforms. Ari Paparo, founder and CEO of Marketecture Media, argued on July 21 that agentic AI poses multiple existential threats to DSPs. The modern DSP represents "one of the most complex categories of software ever invented," according to Paparo, but technological shifts threaten this established model through AI-driven alternatives.

Amazon introduced agentic AI capabilities across its seller platform on September 17. The system transformed Seller Assistant from a question-answering tool into an autonomous agent that monitors accounts, optimizes inventory, and manages advertising campaigns around the clock.

Trust remains implementation barrier

Misha Williams, chief operating officer at GWI, identified trust as a roadblock for marketers when it comes to artificial intelligence. "There remains a hesitancy to fully adopt," Williams told The Current. "What we need is for these agents to have real, human grounded insights at their disposal, so that marketers feel emboldened to experiment."

Despite reservations, Root confirmed that Permutive plans to engage with AdCP because "it's important to shape the direction of where the protocol goes." Multiple companies expressed similar sentiments about participating in development discussions regardless of current implementation plans.

Kahan from Brainlabs articulated this perspective. "As an agency, we have a responsibility to ride the wave and be part of those discussions," he said. "If we aren't, then we lose our voice. We have to vocalize what we need for our clients."

The protocol launched with version 2.0.0 on October 15, introducing what its creators describe as production-ready capabilities for modern advertising workflows. O'Kelley, who founded AppNexus in 2007 and served as chief executive officer until its acquisition by AT&T in 2018, leads the protocol development.

The protocol's pricing model distinguishes between fixed-price inventory and auction-based purchasing. Fixed-price products specify cost-per-thousand-impression rates and minimum spend requirements. Auction products accept bid prices from buyers competing in real-time auctions. Some products support both pricing models, allowing buyers to select their preferred purchasing method.

Kahan noted that at least 20 companies are involved in the consortium. "There's always the risk that the ones making the rules are the ones with their finger on the scale," he said, explaining the importance of agency participation in shaping standards.

Platform providers respond

Google unveiled comprehensive AI advertising tools at Think Week 2025 on September 10, introducing three AI-powered advisory systems described as "agentic capabilities." Your Ads Advisor operates as "an always-on expert for comprehensive help in Google Ads" that learns account structures and business objectives to suggest campaign optimizations.

The company also released an open-source Model Context Protocol server on October 7 enabling Large Language Models to connect with Google Ads API for read-only reporting and diagnostics through natural language queries.

StackAdapt launched Ivy on July 9, an artificial intelligence-powered assistant integrated directly into its programmatic advertising platform. The Toronto-based company positioned the tool as designed to help agencies and brands accelerate decision-making throughout campaign management processes. According to the announcement, Ivy processed more than 1,700 in-platform messages during the 30 days leading up to its launch.

Smartclip introduced Sidekicks on July 15, an agentic AI platform combining media intelligence with process automation. The platform delivers specialized AI agents that assist teams while maintaining organizational knowledge security. RTL Deutschland and Ad Alliance are evaluating pilot integration opportunities for AI-driven automation in campaign management and cross-media planning.

Implementation challenges persist despite growing investment. Gartner predicted on June 25 that over 40 percent of agentic AI projects will be canceled by the end of 2027 due to escalating costs, unclear business value, and inadequate risk controls. The research firm's January 2025 poll of 3,412 webinar attendees revealed uneven investment patterns, with 19 percent reporting significant investments in agentic systems while 42 percent made conservative investments.

The privacy implications favor agentic approaches according to some analysis. AI agents present "less problematic" solutions for privacy and data movement compared to traditional DSP models that require extensive data sharing across multiple parties and platforms.

Why this matters

The marketing community faces fundamental questions about whether existing infrastructure can accommodate autonomous AI systems or whether new protocols like AdCP are necessary. The debate extends beyond technical specifications to strategic considerations about transparency, trust, and control in automated advertising.

For agencies and brands, the proliferation of competing standards creates uncertainty about which systems will gain adoption. The risk of backing the wrong protocol could result in wasted integration efforts and incompatibility with future industry practices.

The tension between innovation advocates and skeptics reflects broader concerns about automation's impact on oversight and accountability. While efficiency gains from agentic systems appear substantial, questions remain about how marketers will maintain control over spending decisions and ensure alignment with brand objectives.

Platform providers including Google, Amazon, and Meta are building agentic capabilities into their existing systems rather than adopting external protocols. This approach could leave open web publishers at a disadvantage unless standardized protocols like AdCP gain sufficient traction to enable competitive automation.

The six founding members represent different segments of the advertising technology ecosystem. PubMatic operates as a supply-side platform connecting publishers with buyers. Scope3 previously focused on carbon emissions measurement for digital advertising before pivoting to agentic advertising. Swivel delivers audience targeting and analytics capabilities. Triton specializes in audio advertising technology for streaming and podcasting. Optable focuses on privacy-first data collaboration. Yahoo brings established relationships with premium publishers and buyers.

Timeline

• January 2024: McKinsey documents $1.1 billion equity investment in agentic AI during 2024
• May 14, 2025: Microsoft announces discontinuation of Microsoft Invest (Xandr) effective February 28, 2026
• July 9, 2025: StackAdapt launches Ivy AI assistant for programmatic advertising
• July 15, 2025: Smartclip introduces Sidekicks agentic AI platform
• July 21, 2025: Ari Paparo publishes analysis warning agentic AI threatens DSP business models
• July 27, 2025: McKinsey Technology Trends Outlook 2025 identifies agentic AI as frontier technology
• September 10, 2025: Google unveils agentic capabilities at Think Week 2025
• September 10, 2025: Adobe launches Experience Platform Agent Orchestrator
• September 17, 2025: Amazon introduces agentic AI across seller platform
• October 1, 2025: LiveRamp debuts agentic orchestration capabilities
• October 7, 2025: Google releases open-source MCP server for Ads API integration
• October 15, 2025: Ad Context Protocol launches with six founding members
•  October 29, 2025: ANA Masters of Marketing conference features agentic AI discussions

Summary

Who: Six advertising technology companies—Scope3, Yahoo, PubMatic, Swivel, Triton, and Optable—founded Ad Context Protocol. Brian O'Kelley, who previously founded AppNexus, leads the protocol development. Industry observers including Anthony Katsur from IAB Tech Lab, Augustine Fou, Joe Root from Permutive, and Ben Kahan from Brainlabs responded to the announcement.

What: Ad Context Protocol is an open-source technical standard built on Anthropic's Model Context Protocol. The framework enables artificial intelligence agents to communicate with each other and perform advertising tasks including inventory discovery, price comparison, and campaign activation across multiple platforms without requiring custom integration work. The protocol provides nine core tasks covering the complete advertising lifecycle and launched with version 2.0.0.

When: The protocol launched on October 15, 2025. Development followed months of growing industry interest in agentic AI, including McKinsey's July 2025 research showing $1.1 billion in equity investment during 2024 and 985 percent increase in related job postings. The announcement came during the same month as the ANA Masters of Marketing conference where agentic AI featured prominently in discussions.

Where: The protocol targets the open web advertising ecosystem. It addresses fragmentation issues affecting media buyers and agencies that currently develop separate integration capabilities for Google Ad Manager, connected television platforms, digital out-of-home systems, and other advertising channels. The founding members represent different segments including supply-side platforms, audio advertising technology, audience targeting, and data collaboration.

Why: Supporters argue the advertising industry is entering an agentic era requiring standardization across fragmented systems. Without unified protocols, each AI agent would need custom integration work for every advertising platform. However, skeptics question whether another protocol is necessary when existing standards like Agent2Agent and Model Context Protocol already exist. Critics also warn the initiative could distract from foundational performance problems and transparency concerns while adding complexity rather than solving core issues. The debate reflects broader tension about automation's impact on oversight and accountability in digital advertising.