Report comment

Please fill in the form to report an unsuitable comment. Please state which comment is of concern and why. It will be sent to our moderator for review.


Given that the lawyer's job is not to predict the outcome of cases but to present their client's case in the best possible way (usually with limited visibility of the other side's case) in order to influence the outcome, then this should not be of (immediate) concern.

If I was a case handler for the financial ombudsman then I might be rather more worried...

The article does not indicate if the lawyers were experienced in dealing with FCA ombudsman cases (unlikely). If not, then lawyers would tend to default to deciding cases on the law, whereas CaseCrunch would reflect the biases, pre-conceptions, and extrinsic considerations that may also drive ombudsman decisions. Ironically, if the ombudsmen were replaced by such automated systems then the result might be to entrench such biases.

An interesting analysis might be to see if the lawyers disagreed with the ombudsmen in a consistent manner. If so, this begs the question of whether the lawyers were "wrong" or whether the original ombudsman decisions were.

Where this tool would be useful is for ODR / industry ADR schemes where disputes can be framed in a manageable manner. Before we get there though, the systems need to be able to be able to ascertain questions of fact rather than merely apply the law to an agreed set of facts.

Your details