2026 Accelerator Incubator

Objectives

Champions: AP (Associated Press), Al Jazeera, Washington Post, BBC, Channel 4, ITV, Sky, EBU, NBC Universal, SMPTE, Reuters, ITN, Scripps, The Global Creative & Security Community

Participants: Shure, EVS, CUEZ, Moments Lab, Perspective Media Group, Google Cloud, TRINT, Cognizant, Nuvelics, AWS, Octopus, The Weather Company, HyperContent AI, LiveU, Electric Sheep, Fonn Group

News production has spent decades connecting best-of-breed systems so that media, commands and operational metadata can move between tools. But the editorial meaning of a story - what is known, what has been verified, what can be published, what changed, and which decisions have been made - still does not travel with the content.

That gap becomes critical as AI and agentic tools enter production. A newsroom system, MAM, graphics platform, planning tool, publisher and AI service can each hold a different fragment of the same story, while people remain the integration layer carrying context from one system to the next - often through memory, messages and conversations outside the production stack.

This IBC Incubator Smart Stories aimed to create a shared, open data foundation for that missing context without replacing the systems that already own media, rundowns, or playout, and without automating away the points where editorial judgement is required.

The Results

Over six months, the Smart Stories consortium designed and implemented the Story Object Model (SOM), an open standard for representing story context so that people, production systems and intelligent tools can share a common understanding of the same evolving story.

SOM models three core layers: the Story - the evolving editorial account of what is happening; Assets - the media and material collected or created to tell it; and Tellings - the point where an asset meets a specific outlet and audience, including the editorial and compliance decisions that govern that use. This allows the model to reduce repetitive handoffs while preserving the human judgement and decision gates that determine what is appropriate to tell, where, and for which audience - protecting editorial quality and trust as workflows become more automated.

The project also developed a shared data-bus approach for systems to read, write and enrich story context, and explored migration paths for existing infrastructure so established newsroom environments can participate without ripping out current integrations.

The result is a vendor-agnostic foundation intended to make best-of-breed production ecosystems more interoperable and to give AI agents structured context rather than forcing them to infer it. SOM also connects to TAMS so story context references media by source and time range without owning it.

Key Goals & Achievements

• Defined SOM as an open, industry-oriented model for story context interoperability. storyobjectmodel.com
• Created a three-layer Story / Asset / Telling model that separates the evolving story, the content used to tell it, and the editorial decision at the point of distribution.
• Encoded editorial state, provenance and attribution, lifecycle, compliance, audience/outlet context and decision history as structured data that tools can consume.
• Designed the architecture to preserve human editorial control by making decision points explicit and attributable rather than removing them as workflow friction is reduced.
• Demonstrated a multi-vendor approach in which different tools and AI services can consume and enrich the same story context without requiring a single monolithic platform.
• Separated the burden of carrying context between systems from the judgement itself, creating a path for automation to handle more coordination while editorial responsibility remains visible and human-led.
• Established a foundation for agentic workflows in which deterministic story state can be shared reliably, and AI systems can act with richer, traceable context.

Why This Matters for the Media Industry

The significance extends beyond a single AI workflow. The media industry has standards and integrations for moving media and operational instructions, but it has never had a common model for moving editorial meaning. Smart Stories proposes making context infrastructure.

If adopted broadly, that shared layer could reduce bespoke point-to-point integration, make best-of-breed systems work better together, improve traceability and auditability, and let AI agents work from explicit editorial state rather than guesswork. It also creates a practical way to relieve skilled newsroom and production teams of the repetitive work of carrying context between systems while keeping the judgement itself - the decisions about accuracy, framing, compliance and audience - with people.

Results Showcased at IBC2026

At IBC2026, the consortium presented the Smart Stories work on the Future Tech Stage through two realistic newsroom scenarios: a storm that becomes a hurricane, including a report that could not initially be confirmed, and a Downing Street statement whose outcome remained unknown until the Prime Minister spoke. Across the demonstrations, solutions from 13 different companies read and updated the same story context using infrastructure from two different cloud providers, showing how provenance, accuracy, synchronization and human approval could travel across the workflow without a central orchestrator.

View the Project's Final Showcase Deck

What's Next?

The consortium’s next focus is to take SOM from an IBC incubator proof point toward an open industry standard and wider implementation. Priorities include refining the specification, validating conformance and governance approaches, expanding integrations and real-world newsroom use cases, refining how editorial policies and decision gates are represented, and creating practical adoption paths for both modern API-based systems and established broadcast infrastructure. storyobjectmodel.com