Swivel hires Samsung, Magnite veterans to scale agentic CTV operations

The transition from experimental AI to 'agentic execution' marks a shift in CTV ad operations toward full automation of sales and yield workflows. By hiring executives from Samsung and Magnite, Swivel is signaling that agentic systems are ready for enterprise-grade deployment outside of walled gardens. This addresses the rising operational complexity of managing high-volume programmatic and direct demand across fragmented streaming platforms. As human teams shift toward strategic guardrail management rather than manual execution, the industry must now reconcile competing automation standards. Watch for the rollout of secondary AdCP modules in late 2026 to see if agent-to-agent transactions become a standard alternative to traditional OpenRTB auctions.

The hiring of Samsung and Magnite veterans by Swivel comes as the industry formalizes the infrastructure for AI-driven trading. In October 2025, a consortium including Yahoo, PubMatic, and Swivel launched the Ad Context Protocol (AdCP), an open standard built on Anthropic's Model Context Protocol. Per AdExchanger and Streaming Media, this protocol aims to standardize how AI agents—capable of negotiating and executing media buys without human intermediaries—communicate across the supply chain, mirroring the role OpenRTB played for programmatic bidding.

Enterprise adoption of these systems accelerated in early 2026. Per IAB Tech Lab in January 2026, the organization released its own 'Agentic Roadmap' to layer automation into existing standards like VAST and OpenRTB. This has led to high-profile deployments, such as NBCUniversal using agentic AI to automate media buying for live football playoff games in Q1 2026. While competitors like the IAB focus on the Agentic RTB Framework (ARTF), Swivel’s expansion underscores a distinct push toward 'seller agents' that provide publishers with natural language tools to manage inventory.

However, the scaling of agentic operations faces significant headwinds regarding cost and risk. Per Gartner in late 2025, an estimated 40% of agentic AI projects are projected to be canceled by 2027 due to unclear business value or the inability of teams to explain machine-led decisions. Swivel’s focus on hiring operational experts like Melaragno suggests that the next phase of the market will be defined by the 'operational design' of these workflows—training teams to monitor model performance and maintain human oversight within automated guardrails. Integration remains a key hurdle as publishers navigate fragmented data environments and a dual-track ecosystem of legacy programmatic pipes and emerging agent-to-agent protocols.