Research Finds: Nearly 10% of AI Chatbot Users Engage in Conversations with Pornographic Themes

aibase

站长之家

Published in AI News · 1 minute read · Jul 17, 2025

Translation: Latest research indicates that nearly 10% of AI chatbot users misuse the technology for explicit conversations. The research team analyzed 100,000 chatbot dialogues and found that approximately 10% involved adult themes. The study aims to uncover the true interactions between people and chatbots, enhancing the safety of chatbots in real-world applications. Researchers categorized the misuse into three types: obscene narratives, sexual fantasies and role-playing, as well as discussions of harmful behaviors.

Related AI News

Baidu Launches a Series of AI Products Including AI Volunteering Assistant and College Entrance Examination Big Data

Baidu Launches a Series of AI Products Including AI Volunteering Assistant and College Entrance Examination Big Data

With only 8 days left before the 2025 college entrance exam, candidates face immense pressure and challenges. To help candidates review more effectively during the final sprint phase, Baidu has released a series of innovative AI tools. These tools aim to provide comprehensive support from reviewing for exams to filling out applications, helping each candidate better cope with the upcoming exam. First, Baidu launched the 'Frequent Exam Topic Library for College Entrance Examination', which gathers core topics from the past three years. Candidates just need to open the Baidu App, search for 'college entrance examination', and they can easily access all the topics from various provinces.

OpenAI Makes a Big Move! Image Generation API Upgrade: Real-Time Streaming Preview + Multi-Round Editing MCP Enables Boundless Creation!

OpenAI Makes a Big Move! Image Generation API Upgrade: Real-Time Streaming Preview + Multi-Round Editing MCP Enables Boundless Creation!

OpenAI announced significant updates to its image generation API (Responses API), adding new features such as streaming transmission, multi-round editing, and integration with Model Context Protocol (MCP) tools and real-time web data search. These new features not only improve the efficiency of image generation but also provide developers with more flexible creation methods, marking a new stage for AI-driven visual content creation. AIbase has compiled the core highlights of this update and its far-reaching impact on the industry.

Meta Team Research Finds: Simplifying Reasoning Chains Can Significantly Enhance AI Accuracy

Meta Team Research Finds: Simplifying Reasoning Chains Can Significantly Enhance AI Accuracy

Recently, Meta's FAIR team and researchers from the Hebrew University of Jerusalem jointly released a new study indicating that reducing the reasoning time of large language models can significantly improve their performance in complex reasoning tasks. The research findings show that using shorter reasoning chains, the accuracy of AI models has increased by 34.5%, which challenges some assumptions in the current AI industry. Image source notes: The image was generated by AI and is licensed by Midjourney in this study, the authors point out that prolonged deliberation

Ant Group open-sources Ming-lite-omni: The first open-source multimodal model comparable to GPT-4o

Ant Group open-sources Ming-lite-omni: The first open-source multimodal model comparable to GPT-4o

Tesla plans to launch Robotaxi service in Austin on June 12

Tesla plans to launch Robotaxi service in Austin on June 12

Tesla is about to launch its autonomous taxi service. Sources reveal that Tesla Inc. plans to introduce its long-anticipated Robotaxi autonomous taxi service in Austin on June 12, 2025. This marks another significant step forward in Elon Musk's strategic layout for self-driving cars and artificial intelligence. Background Information: Tesla’s Vision for Autonomous Driving Since its inception, Tesla has been committed to revolutionizing the automotive industry, particularly in the field of autonomous driving technology.

# Musk Attempts to Block OpenAI AI Deal in the Middle East

# Musk Attempts to Block OpenAI AI Deal in the Middle East

Body: In a major recent development, U.S. technology giant OpenAI has reached a cooperation agreement with G42 of the UAE to establish one of the largest artificial intelligence data centers in the world in Abu Dhabi. However, behind this deal lies significant pressure and intervention from Elon Musk. According to relevant reports, Musk has strongly discouraged this cooperation during a call with G42 officials, stating that the agreement would be difficult to obtain approval from the U.S. government unless it includes his company xAI.