Accountancy’s AI data dilemma

Bookkeeping and accountancy is a promising industry for AI-native software.

Jenesys, Briefcase, and Artifact AI have all been built with artificial intelligence (AI) baked in. These apps - plus others not yet in the public domain - promise big things from their native application of AI. 24/7 bookkeeping; continuous close; even daily insights. The potential matches the hype.

There is a catch, though. How useful AI-native software will be for firms depends on how well the AI is trained. The quality and quantity of data fed to an AI training model determines its value for a firm and its clients, and there’s only so far a model can go with what’s publicly available. 

As Briefcase writes, “Foundational models are trained on publicly available data and don't have the specific business knowledge that accountants need.” What's required is "specific data" that allows "superior accuracy and reasoning capabilities" - and providing training models with this "contextual learning" is Briefcase's preferred approach.

So where will AI software vendors get this specific business knowledge? It’s in your client data!