We're lucky to work closely with a number of top-25 Pharma companies and one insight stands out in how they are bringing AI into their organizations: intentionality.

These leaders understand the AI world is moving incredibly quickly, so internal initiatives should be selected carefully and designed to:

• Take advantage of ongoing model & infrastructure improvements
• Deliver tangible solutions to internal teams
• Own sensitive areas (i.e. confidential data, access to internal systems)

Just within these buckets lies a ton of complexity. The scale of these companies is such that the rollout of AI initiatives will be complicated by nuances in product requirements, user needs and AI fluency across teams and geographies.

And when billions of dollars have been spent building proprietary human, biological and knowledge-based datasets, you need to balance the race to unlock that data with genuine privacy & security constraints.

Given all that, it makes sense these orgs look to partner outside of those core competencies. By being intentional about the areas you need to own internally, you avoid spreading teams too thin, and allow individual stakeholders to integrate external tools & datasets that advance their business outcomes.

For example:

• BD, strategy and CI teams partner with Sleuth for AI-native pipeline, deal and trial intelligence
• Research scientists use Edison Scientific's Kosmos to synthesize literature and generate novel hypotheses

In both cases, fit-for-purpose data, insights and outputs are piped directly to internal AI agents & workflows. This interoperability is a huge unlock - the most AI native orgs will know exactly what to own and make everything else easy to integrate.

If your BD, strategy, or CI team is working through the own-vs-partner question, that's the conversation we have most often. Happy to compare notes on where other top-25 organizations have drawn the line.