How are traders on credit desks using AI for CDS work today? Curious for inspo for myself as a junior in the field.

Been thinking a lot about where AI is actually useful in credit/CDS vs where it just sounds good in a demo.

Say you have a credit event/restructuring and you’re trying to figure out what it means for the CDS, which obligations matter, what’s deliverable, what the docs actually say, similar historical cases, etc.

Are you using ChatGPT/claude for any of this today? Or is it still mostly Bloomberg/S&P/ICE + docs + spreadsheets + legal/internal tools old school style?

My feeling is the general models are probably useful for reading/summarizing, but I’d be pretty nervous trusting one on the actual conclusion without checking everything myself.

Also curious if anyone has ever properly benchmarked the different models on real historical CDS cases rather than just testing a few prompts.

Would be interested to hear what people actually on desks are doing, and where AI is useful vs completely useless.

Thanks all// mr B