"NYC's Misleading Chatbot: A Threat to Compliance and Ethics"

NYC's government chatbot, "MyCity," has been found to provide dangerously wrong information about city laws and regulations, including misinformation about housing subsidies, worker pay regulations, and industry-specific information. The chatbot, powered by Microsoft Azure, uses statistical associations to guess at answers without real understanding, leading to inaccuracies. Despite warnings about potential incorrect, harmful, or biased content, the chatbot is being promoted as a way to help business owners navigate government. This highlights the risks of rolling out chatbots before their accuracy and reliability have been fully vetted, as seen in other instances of chatbots providing misleading or inaccurate information.
- NYC's government chatbot is lying about city laws and regulations Ars Technica
- NYC's AI Chatbot Tells Businesses to Break the Law – The Markup The Markup
- New York City's AI chatbot is telling people to break laws and do crimes Quartz
- New York City's Microsoft Chatbot May Tell Small Business Owners It's Okay to Cheat Inc.
- City chatbot fail, Rikers commissary, landlord income THE CITY
Reading Insights
0
12
3 min
vs 4 min read
87%
777 → 99 words
Want the full story? Read the original article
Read on Ars Technica