Date: April 07, 2024
OpenAI might have to pause its ChatGPT releases following a complaint filed by the Center for AI and Digital Policy (CAIDP) with the Federal Trade Commission (FTC).
The nonprofit organization claims that OpenAI is violating the FTC Act by releasing "biased, deceptive" AI models like GPT-4, which could potentially threaten privacy and public safety. The CAIDP wants the FTC to investigate OpenAI and suspend future releases until the models meet the agency's guidelines.
Additionally, the researchers are demanding that OpenAI require independent reviews of their GPT products and services before launching them. CAIDP also hopes that the FTC will create an incident reporting system and formal standards for AI generators.
Critics of ChatGPT and similar models have warned of problematic output, including inaccurate statements, hate speech, and bias. CAIDP points out that users cannot repeat results, and OpenAI itself warns that AI can "reinforce" ideas whether or not they're true.
While upgrades like GPT-4 are more reliable, there's a concern that people may rely on the AI without double-checking its content. It's not guaranteed that the FTC will act on the complaint, but if it does set requirements, it could affect development across the AI industry.
Companies would have to wait for assessments and might face more repercussions if their models fail to meet the Commission's standards. While this could improve accountability, it could also slow down the currently rapid pace of AI development.
Google Threat Intelligence Will Use Gemini For Cybersecurity
Google Threat Intelligence is the latest introduction by Alphabet INC. and will function as the main cybersecurity layer for its service offerings.
Meta May Soon Launch Generative AI Ad Offerings
Meta is planning to expand its AD offerings to facilitate AI-powered tools that automatically create versions of images and align text on them.
Opera’s AI Assistant On Android Can Now Summarize Web Pages
Opera Browser’s in-app AI assistant, Aria, has upgraded to new capabilities, including summarizing the content of web pages on Android.
This Year’s Met Gala Theme Is Apparently Deepfake AI
Celebrity culture is filled with outward themes, and the Met Gala is one of the most prominent platforms to display them. This year, the style is Deepfake.