$CSU and $TOI AI disruption risk is somewhat hard to quantify, as there are so many subsidiaries.

But, @LigCapital_ on X tried doing exactly this. Link to his work below:  

Interesting data shared, companies exhibiting traits of:

System of Record: yes (53%), partial (42%)

Mission Criticality: high (79%), partial (16%)

Compliance moat: high (40%), medium (21%)

Replicability: low (34%), medium (57%)

Domain expertise: high (77%), medium (18%)

Conclusion: Only 10% of subsidiaries in high to very high risk for AI.

Although this will never be perfect, we've never found anyone trying to quantify this before. Kudos for this one