Who, Where, Why?

The National Institute of Standards and Technology (NIST) have been the authority on what is and isnt acceptable in the technology world for as long as I have been working. The news that they plan to release a list of "trustworthy and responsible" AI resources hopefully means that the acceptance and uptake in the use of these powerful tools will start to become more "normal".

Although NIST are based in Gaithersburg, Maryland (USA), a lot of their frameworks and guidance is universally accepted and relevant to technology in every country across the world.

How is this relevant to my cases?

The biggest concern I am presented with when I suggest the use of AI in litigious cases is that the users don't feel comfortable as they do not fully understand how it works. This in turn makes standing behind the work product more difficult compared to how it would have been had the work been done the "old way".

I always do my best to address those concerns and believe that it is the responsibility of the service provider to ensure that the way in which AI is implemented is explained and understood before any work is started. Having a resource available to assist in that process such as the new NIST repository is inevitably going to mean that this is made easier. 

There is a time and a place for AI!

It is worth noting that the resource is only as good as the people who use it. As eDiscovery professionals the onus is on us to ensure we are comfortable with the services we are offering to our clients and just because NIST say something is OK to use, it doesnt mean that we should all rush out and use it on every case.

The right tool should always be explored as an answer to each unique problem - including using our traditional toolkit when that unique problem doesnt need the all singing, all dancing AI solution. 

The best way to build trust with people is to prove you can be trusted to give them an answer that works best for the situation every time you are given the oportunity.