As generative AI moves from experimentation to production use in live discovery matters (Human AI-Assisted Review or HAR), the legal profession faces a question it has been collectively reluctant to answer: is the validation architecture that legitimised Technology-Assisted Review fit for purpose when applied to a technology that works very differently?
A new Legal IT Insider report, out on 9 March, draws on interviews with the leading voices in this area, including Judge Paul Grimm, Professor Maura R. Grossman, Professor Jason R. Baron, litigation and eDiscovery specialist Philip Favro, Jim Sullivan of eDiscovery AI, and senior practitioners at Consilio, Epiq, OpenText, Control Risks, Everlaw, and other major discovery service providers.
It argues that reliance on HAR under just the existing TAR framework is likely to lead to a major discovery failure and judicial sanction. More importantly, the article proposes what that necessary alternative framework looks like.
The report is written by lead analyst, legal technologist and former barrister Neil Cameron.
If you are attending Legalweek New York or BLTF London, be sure to come and find us as we will have limited printed copies available too.
The post From TAR to HAR: Are GenAI discovery tools ready for forensic scrutiny? appeared first on Legal IT Insider.