AI-Written Laws and Legal Work Raise Concerns Over Bias and Accuracy
• A politician in Brazil used ChatGPT to write legislation that got passed without the city council knowing • Upon learning the bill was AI-written, some politicians became angry and want it revoked as a "dangerous precedent" • A judge in the UK and India have also used ChatGPT for legal matters, but experts warn of bias and technical issues • The AI tool struggles with citations and has created fake legal precedents, causing problems in court • It has shown political bias, such as being more tolerant of hate speech against Republicans and once refusing to write a NY Post-style article