01:00
2026-07-17
smarterarticles.co.uk
artificial-intelligence
The Information Underclass: How AI Delivers Worse Answers to Vulnerable Users
A February 2026 study by MIT's Center for Constructive Communication found that three leading AI chatbots—OpenAI's GPT-4, Anthropic's Claude 3 Opus, and Meta's Llama 3—deliver significantly worse answ…