Synthetic Narrative: The Ethics of Generative Storytelling in Institutional Portfolios

Every serious media investor is now being asked the same two questions, in the same order. First: how much of your portfolio's production spend can generative systems absorb? Second: what happens to the value of your intellectual property when they can absorb even more?

Our position, formalised in our H2 2026 investment memorandum, is that these are not the same conversation. Generative tools are, undeniably, compressing costs in visual effects, pre-visualisation, dubbing, and post-production. Portfolio companies in our infrastructure strategy report gross-margin improvements between eleven and eighteen percent on projects that have integrated production-grade generative pipelines. That is a material tailwind.

The intellectual property question is more subtle. The economic worth of a story is not the pixels it renders into; it is the exclusive right to render them. As generative systems make execution cheaper, the scarce asset becomes the underlying rights — the novel, the true event, the recognisable character, the actor's likeness. This is a re-pricing, not a devaluation. It rewards holders of clean, well-documented, contractually defensible IP libraries. It punishes holders of derivative, ambiguously-sourced material.

We have accordingly tightened three internal standards. First, provenance: every acquisition in our IP Library strategy now requires a full chain-of-title audit that assumes adversarial future litigation. Second, consent: all talent agreements in productions we finance now contain explicit, opt-in language governing synthetic likeness and voice usage, negotiated deal-by-deal. Third, disclosure: portfolio companies deploying generative tools in principal creative decisions — as opposed to post-production efficiency — are required to disclose that use to us, to their distribution partners, and where relevant to their audiences.

None of this is a hedge against the technology. It is a bet that the market will, over the next twenty-four months, sharply differentiate between institutions that have thought carefully about these boundaries and those that have not. We intend to be firmly in the former category.

The dispatch published in H1 2027 will return to this theme with the first hard performance data from our synthetically-augmented productions. Until then, our guidance to partners remains unchanged: invest in stories worth protecting, and protect them like they are worth investing in.