What is your approach to protect intellectual property artifacts ? Do you use a DLP in-line or into the workstations? Have you implemented a DRM ?  Do you use virtual machines to segregate access to data?

The only place that was close to controlling their data was the US Military. They spent vast sums of money building redundant networks with the data isolated to each network. Even then, it was not clean, just bigger piles of data in isolated pools.
In the corporate world, companies do not put a priority on classifying data. The older data is a huge black hole of data that is mostly un-classified. If it doesn't have a classification, how do you protect it? Think of the labor hours needed to sort these terrabytes of data and the lack of line items when setting the yearly budget.
Using Windows desktops and servers allows you to use Microsoft tools to monitor your data. You can trace the documents movements, you can prevent access. Again, there is an expense and it takes a lot of time to implement.
Most companies are focused on delivering the next project, it is extremely rare to see a company start doing the drudgework of data sanitation. However, if you start cleaning your data, moving it into new controlled locations with then you can use these clean groupings to add an AI element on your internal data safely. This should be viewed as an opportunity to clean and tag the data, get it under control, and then use the data for internal AI agents. If you feed in sanitized and tagged data, you may then carry that classification onto your AI agent. For example, public data is one agent, finance data is internal use only. That way you reduce the risk of leaking your financials or PII into the wild.

Data is king! So encryption, end-point and network DLPs, and EDR - and move towards ZT-access management. Enclaves work great when you have many many data layers that need segregation. No one strategy serves as the silver bullet in my experience.