SyncAI.news, a Varaisys broadcasting
When AI Says It Is Compliant, What Exactly Does It Know?
DP

Darko Pavic, Forbes Councils Member

· 1 min read

World NewsForbes: Innovation

When AI Says It Is Compliant, What Exactly Does It Know?

Darko Pavic, Founder & CEO of Fiscal Solutions, Retail Technology & Global Fiscalization Expert with 28+ Years of Experience.

​Imagine a typical situation at any global retailer: an international rollout of a new POS solution. After months of preparation, the solution and the team are ready. Part of the preparation involved consulting with AI about legal requirements in all targeted countries.

Nowadays, companies and teams are encouraged to do exactly that. Even on the transaction level, teams are testing POS workflows and asking AI, “Is this transaction compliant with local regulations and laws?”

AI reviews the transaction, analyzes applicable laws, answers “yes” and explains everything. After going live, the retailer discovers that the AI’s answer was wrong. AI understood all the local rules that it found perfectly. Unfortunately, it didn’t know about one decisive technical requirement.

The answer didn’t show the gap, and the retailer wasn’t presented with a lie. AI simply based its answer on the information that it had. In this process, the retailer got something very dangerous: a confident conclusion and statement based on incomplete legal coverage.

When AI says something is compliant, can you trust the answer?

Moving Beyond The Transaction

I have now been working for more than two decades with fiscal regulations in retail and software implementations across the world. What I’ve learned is that failures are mostly not based on huge misunderstandings of laws. Very often, they are caused by small gaps in the coverage of legal requirements: the wrong version of a legal or technical specification, a missing official interpretation or a small exception hidden in another document or even another regulation.

Clarifications and exceptions are often the reason why legal compliance fails in a global environment. And now, as AI is taking the lead, these old problems are simply being amplified.

What Is AI Good For?​

The Future Of AI Governance

• What exact decision am I making?

Original source

This story was published by Forbes: Innovation and written by Darko Pavic, Forbes Councils Member. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on forbes.com

Similar News