In times of immense and rapid change, there is nothing more valuable than convening a group of experts to discuss the defining topics of the day. Plural’s Value Creation in Information Services summit (VCIS) 2026 brought together over 220 C-level leaders and investors from across the data and information sector at BAFTA in London, spanning every vertical from finance, energy, legal to insurance, media, maritime and more. The themes of the day were clear and very consistent:
- A clear gap is emerging between operators’ views of their resilience and AI defensibility, and the perception of the wider market
- Proprietary data means more than owning non-public data: the most trusted outcome-oriented datasets combine human expertise and sourcing, proprietary methodologies and derived datasets (e.g. scores, indices, rankings, forecasts)
- New distribution and access modes, e.g. MCPs, create potential for significant usage pricing uplift, but require new and sophisticated approaches to licensing and commercial models
- Human expertise and relationships are more important than ever as sources of data, expertise and as markers of trust
Operating in this new environment, closing the perception gap, and harnessing the genuine opportunities of AI will require a leadership-led culture of continuous, customer-focused innovation and transformation, and a new approach to external communication. For those that are slow to change, a significant value gap risks emerging.
A gap is emerging: closing it starts with customers
The results of Plural’s survey of senior leaders across the sector highlighted the gap:
- Whilst 76% of senior data and information executives agree that AI is a net positive for established data and information businesses, 88% believe others will not survive the transition in their current form
- At the same time, only 42% feel confident in their ability to explain their AI resilience to investors, and only 7% feel they are doing a good job of communicating their differentiated value vs AI alternatives to customers.
There are two forces driving this:
- A perceived innovation deficit: an external and investor perception that alternatives can do things better, faster and cheaper, and that established businesses are not doing enough to keep pace (whether true or not).
- A communication and narrative gap: operators are positive about their own businesses, concerned about others, while the market is concerned overall, and the narrative to bridge those views is not.
Closing both starts with focused innovation led by real proximity to customers, getting under the skin of the business outcomes they are trying to achieve, the tools they are using and the alternatives they have, and creating solutions to these.
Proprietary data must be hard to replicate, not just hard to find
Data that was difficult to aggregate a few years ago is now accessible to well-resourced competitors, and the instinctive answer of ‘we have proprietary data’ is not the only foundation for defensibility. Defensibility is broader and more nuanced than simply owning data: what remains genuinely proprietary includes data sourced through exclusive relationships, derived scores and forecasts that require genuine methodological depth to build and to trust, and contextual insight that cannot be found in the public domain.
AI data licensing can command significant price premiums
It was clear from commercial conversations that alternative modes of data access will command very different price levels and require different pricing models:
- A standard user interface at the baseline, an API feed at approximately 3 x that, large-scale data integrations at 4-5 x and data licensed via an MCP directly into a customer’s own AI environment at a further five x on top of that.
- Token-based pricing is beginning to appear alongside traditional licence models, with a lighter “conversational” tier emerging for customers who simply need answers to specific questions, with potential to expand the market and usage.
However, there are also risks: when your data sits inside someone else’s model, your relationship with the end customer becomes more distant, making licensing terms and redistribution discipline as important as the product itself.
Human expertise is gaining in value
As AI makes data easier to generate and collect, things that cannot be replicated are becoming more valuable. The counterintuitive insight of the day was that human expertise is gaining value. Journalists, analysts and subject matter experts who source information through exclusive, trusted relationships and on-the-ground presence are now a primary source of genuinely proprietary signals, requiring deliberate investment in incentive structures, workflows, and in how their knowledge is captured and converted into structured intelligence.
AI also enables the conversion of unstructured content and event intelligence into structured, forward-looking data at scale. A content business can become a data business, not by abandoning editorial judgement but by applying it more systematically and extracting its signals more efficiently, with journalists and analysts measured on the data points they source as well as the content they create.
For any business sitting on years of unstructured content, it is a real and largely untapped opportunity.
