{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/50c816ad745a46ab85bec9fbada04587\" frameborder=\"0\" width=\"1920\" height=\"1440\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1440,"width":1920,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1440,"thumbnail_width":1920,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/50c816ad745a46ab85bec9fbada04587-f2606c0bc0b5335d.gif","duration":1480.926,"title":"HOW LAWYERS CAN USE GENERATIVE AI TO FIND HALLUCINATIONS IN THEIR WORK","description":"I discussed the growing problem of AI hallucinations in legal work and why courts are sanctioning lawyers for them. I cited examples like CHA sanctioned nearly $60,000 for using ChatGPT in a court case, and there were 1,031 documented hallucination cases in February, with $86,000 being the largest sanction. I explained three causes, sycophancy, autocomplete on steroids, and data cutoffs and gaps. I recommended human review as the most important safeguard, and then using generative AI to check documents you already drafted. No specific action was requested from viewers, but I implied you should verify every word and citation and use AI review to speed up hallucination checking."}