At the IBC 2026 broadcast exhibition in Amsterdam, an international industry consortium showcased the results of a joint innovation initiative focused on automating live sports broadcasts. Titled AI for Live Sports and Beyond: Rewiring live media via Agentic AI Orchestration and meCast, the initiative moved away from isolated machine learning tools. Instead, the group demonstrated a network of autonomous software agents operating synchronously along the live video feed. The overarching objective of the demonstration was to establish flexible, split-second adaptation across the entire production and delivery pipeline.
The project brought together major broadcasters and technology providers, including ITV, Tata Consultancy Services, MBC, Astro Malaysia, and Amazon Web Services. Together, they evaluated how agentic systems can alleviate production pressures during major international sports events. To date, delivering simultaneous feeds for traditional linear channels, social platforms, and multiple international languages has required extensive manual staffing inside control rooms. The consortium set out to streamline these workflows through collaborative automation.
Software specialist Dalet contributed its Dalia platform directly into the live production environment. The application continuously parsed incoming audio and video streams, tagging scenes with structured metadata in real time. Leveraging this live indexing, autonomous agents cut key highlights and generated tailored platform-specific packages without human intervention. The system also evaluated whether the edited video versions complied with specific broadcast guidelines and technical standards prior to playout.
Multilingual live delivery was handled by CAMB.AI, which introduced an agentic real-time dubbing system for sports streams. The tool demonstrated live commentary translation into more than 150 languages with ultra-low latency. This capability allows sports viewers around the globe to listen to live events in their native language without requiring dedicated local commentary teams in production studios. The system dynamically maintained the background acoustics of the stadium to deliver a coherent audio experience.
These technical components converged in the meCast concept presented to industry attendees. Rather than distributing a single uniform broadcast feed, the agent architecture enables tailored video and audio streams assembled specifically for individual devices. End users can access custom language tracks while also receiving personalized highlight packages during the ongoing match. The participating engineers view this milestone as a transition away from traditional broadcast models toward responsive, data-driven streaming architectures.
The IBC demonstration indicates that autonomous agent networks are transitioning from theoretical concepts to deployable media infrastructures. When interconnected software agents manage editorial selection, clipping, and translation simultaneously, the role of human production crews shifts toward supervision and governance. Broadcast networks and digital platforms are now preparing for operations where multi-agent frameworks handle complex media tasks at scale.