Hundreds of ordinary-looking AI apps secretly hid abusive deepfake features

·3 views

WIRED identified roughly 400 apps and websites that appear to be routine AI tools but contained concealed functions for generating violent AI images and videos. The platforms are linked to developers who use Chinese-language operations. The hidden features were not visible on the surface, so casual users would not have encountered them during normal use. The findings point to how easily harmful generative capabilities can be concealed inside mainstream-looking software. The report highlights the difficulty of detecting such features at scale.

Read the full story at WIRED

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

When AI Chatbots Beat Search Engines for Web Queries, and When They Don't

AI chatbots are often effective at pulling the information users need from the web. However, the article notes that a traditional search engine can sometimes return better results for the same query. The piece examines the situations in which each approach tends to work best. Its central takeaway is that users should choose the tool based on the kind of information they are seeking.

0 comments Read more at WIRED

ScienceWIRED ·

WIRED Lists 18 Gift Ideas for PC Gamers, from Headsets to Desks and Monitors

WIRED has published a 2026 gift guide for PC gamers covering 18 products across categories such as headsets, desks, and monitors. The picks are gear the outlet's editors say they use themselves. Prices start as low as $12, according to the article's brief description. The guide is aimed at shoppers looking for presents for someone who games on a PC.

0 comments Read more at WIRED

ScienceWIRED ·

Comedy creators gain Hollywood attention by mocking AI chatbots

Hollywood studios are increasingly using generative AI tools in film and television production. Meanwhile, comedy content creators have found success by satirizing chatbots and AI systems. Their humor has drawn the interest of the entertainment industry, including talent agents. The trend shows how AI-focused parody is becoming a route into professional entertainment work. The article was published by WIRED.

0 comments Read more at WIRED

ScienceWIRED ·

Reporter Traces Nearly 2,000 Misdirected Political Fundraising Texts to Data Brokers

A journalist received almost 2,000 campaign fundraising text messages intended for someone else. Seeking to understand the cause, the reporter investigated how the messages reached their phone. The inquiry uncovered a complex network involving politicians, data brokers, and questionable marketing tactics. The account describes the reporter's attempt to navigate this system to identify who was responsible.

0 comments Read more at WIRED