Is anyone using Generative AI to automatically execute the normally laborious process of creating a data dictionary, establishing data governance, forming questions and hypothesis in a data-driven way? My hypothesis is that someone must be pointing a GenAI engine based on a language model (ex ChatGPT-4) at a data store and having it sort all of that out. I'd be grateful for any pointers or links to firms (even startups) who might be doing that. 

1.5k viewscircle icon1 Upvotecircle icon2 Comments
Sort by:
Chief Information Officer in Software3 years ago

I have to agree with Mark Arrendondo. How much do you trust something like ChatGPT-4 to protect and secure any sensitive data or IP that might be contained within those data streams? You have to be concerned anytime you are exposing data through 3rd Party APIs. 

Definitely a laudable goal to help expedite and cost cut processes, but I would definitely review what I would be exposing to externally hosted AI engines.  I would also make sure I had liability contracts in place.

Lightbulb on1
Director of IT in Healthcare and Biotech3 years ago

We aren't doing this yet because we don't want to expose proprietary information to  a GenAI model at this time.  I think the idea has great merit as it is a pain to do that.

Lightbulb on2

Content you might like

We are not able to do this in real time as of now7%

We do quarterly manual reviews or yearly interviews with the 3rd parties25%

We depend on the 3rd parties to let us know out of band if they feel they have violated the DPA, SCC31%

We have a dedicated personnel that monitors these data transfers using homegrown tools18%

We use outbound firewalls, CDNS, load balancers, API gateways - to get some visibility into these issues, but don't have a complete workflow in place9%

We have built homegrown custom solutions to address this4%

We use commercial services to answer if there is a difference between expected DPA, SCC behavior and reality in terms of data being sent to 3rd parties3%

View Results

Yes74%

No25%