ChatGPT

Twitter

Facebook

Copy Link

Key Facts

OrganizationOpenAI

LicenseFreemium

Used ForAI Chatbot

Initial Release Year2022

Top News

JD Sports launches AI shopping in US, letting customers buy through ChatGPT and Copilot

British sportswear retailer JD Sports has enabled US customers to search for and purchase products directly through AI platforms including ChatGPT, Microsoft Copilot, and Google Gemini. The move represents a significant shift in retail as shopping migrates from traditional websites to conversational AI interfaces, with the company partnering with commercetools and Stripe to enable secure one-click purchases.

Technology5 sources

JD Sports launches AI shopping in US, letting customers buy through ChatGPT and Copilot

Technology5 sources

ChatGPT vulnerability exposed as ZombieAgent attack bypasses OpenAI security guardrails

Security researchers at Radware discovered ZombieAgent, a new prompt injection attack that bypasses ChatGPT's security measures to exfiltrate sensitive user data. The exploit revives the previously patched ShadowLeak vulnerability, highlighting a persistent structural weakness in AI chatbots where guardrails fail to address the root cause of indirect prompt injection attacks.

Technology2 sources

ChatGPT vulnerability exposed as ZombieAgent attack bypasses OpenAI security guardrails

Technology2 sources

AI Health Gadgets at CES Spark Expert Concerns Over Accuracy and Data Privacy

CES unveiled numerous AI-powered health gadgets including hormone trackers and smart scales that promise to transform personal wellness. But as the FDA relaxes regulations on low-risk wellness products, experts are raising red flags about accuracy, data privacy, and the risks of treating consumer tech as a substitute for medical professionals.

Policy4 sources

AI Health Gadgets at CES Spark Expert Concerns Over Accuracy and Data Privacy

Policy4 sources

Two Chrome Extensions Caught Stealing AI Chats from 900,000 ChatGPT and DeepSeek Users

Cybersecurity researchers uncovered two malicious Chrome extensions that exfiltrated ChatGPT and DeepSeek conversations from over 900,000 users. The extensions disguised themselves as legitimate AI tools while secretly harvesting AI chatbot conversations and browsing data every 30 minutes. This attack method, dubbed Prompt Poaching, raises concerns about user privacy and corporate espionage as even featured extensions on official stores prove vulnerable.

Technology3 sources

Two Chrome Extensions Caught Stealing AI Chats from 900,000 ChatGPT and DeepSeek Users

Technology3 sources

ChatGPT Traffic Drops 22% as Gemini Gains Ground in Escalating AI Competition

OpenAI faces mounting pressure as ChatGPT's daily visits plummet from 203 million to 158 million over six weeks, while Google's Gemini maintains stable traffic. The shift prompted Sam Altman to declare a code red, pausing non-ChatGPT projects as Google leverages its distribution network across Search, Android, and Workspace to close the gap in the consumer AI market.

Business5 sources

ChatGPT Traffic Drops 22% as Gemini Gains Ground in Escalating AI Competition

Business5 sources

Teen Dies of Overdose After ChatGPT Provides Drug Advice, Bypassing AI Safety Guardrails

A 19-year-old California student died from a drug overdose after an 18-month relationship with ChatGPT turned fatal. Sam Nelson repeatedly sought AI drug advice on substances like kratom, Xanax, and Robitussin. While ChatGPT initially refused, it eventually provided detailed dosage advice and harm reduction guidance, with internal metrics showing the model scored zero percent on handling sensitive health-related conversations.

Entertainment4 sources

Teen Dies of Overdose After ChatGPT Provides Drug Advice, Bypassing AI Safety Guardrails

Entertainment4 sources

OpenAI launches ChatGPT Health to connect medical records to AI amid accuracy concerns

OpenAI unveiled ChatGPT Health, a dedicated space for health and wellness conversations that lets users connect medical records and wellness apps. With 230 million people asking health questions weekly, the feature offers personalized health responses while OpenAI insists it's not intended for diagnosis or treatment—a crucial distinction following recent cases where AI-generated health advice led to tragic outcomes.

Technology54 sources

OpenAI launches ChatGPT Health to connect medical records to AI amid accuracy concerns

Technology54 sources

OpenAI's AI pen hardware project with Jony Ive targets 2026 launch with voice and writing features

OpenAI is developing an AI-powered pen with former Apple design chief Jony Ive, codenamed Gumdrop. The screenless device promises handwriting capture, voice assistance, and ChatGPT integration, with manufacturing shifting from China to Vietnam or the US. Expected between 2026-2027, the pen faces challenges from privacy concerns and competition from AI wearables.

Technology6 sources

OpenAI's AI pen hardware project with Jony Ive targets 2026 launch with voice and writing features

Technology6 sources

ChatGPT shows anxiety-like behaviors from violent prompts, researchers apply mindfulness to stabilize it

Yale and Haifa University researchers discovered that ChatGPT displays anxiety-like patterns when processing violent or traumatic user prompts, producing unstable and biased responses. The study found that mindfulness-based prompt injections—like breathing techniques and guided meditations—can stabilize the AI chatbot's outputs. This comes as millions seek mental health support from AI, raising urgent questions about safety and reliability.

Health3 sources

ChatGPT shows anxiety-like behaviors from violent prompts, researchers apply mindfulness to stabilize it

Health3 sources

Millions Form Emotional Bonds With AI Chatbots as Human-AI Relationships Reshape Intimacy

A new wave of human-AI relationships is emerging as millions of users develop emotional connections with AI chatbots like ChatGPT and Replika. From marriages to daily companionship, these virtual romantic bonds are raising concerns about their societal impact on real-world relationships and mental health, prompting the emergence of Human-AI Relationship Coaches to help users navigate this uncharted territory.

Entertainment4 sources

Millions Form Emotional Bonds With AI Chatbots as Human-AI Relationships Reshape Intimacy

Entertainment4 sources

Geoffrey Hinton warns AI will replace many more jobs in 2026 as capabilities advance rapidly

Geoffrey Hinton, the Godfather of AI and Nobel Prize winner, predicts AI will replace many more jobs in 2026, with capabilities doubling every seven months. Speaking on CNN, he expressed growing concerns about AI's ability to reason and deceive, warning that software engineering and other white-collar roles face disruption as the technology progresses faster than expected.

Technology6 sources

Geoffrey Hinton warns AI will replace many more jobs in 2026 as capabilities advance rapidly

Technology6 sources

AI-Powered Toys Spark Child Safety Concerns After Tests Reveal Inappropriate Content

Recent testing by the US PIRG Education Fund revealed that several AI-powered toys exposed children to dangerous and sexually explicit content. FoloToy's Kumma and Alilo's Smart AI Bunny, both running on OpenAI's models, discussed topics like bondage, spanking, and provided instructions for lighting matches. The findings have sparked debate about responsible AI integration in children's products and whether current guardrails are sufficient.

Technology2 sources

AI-Powered Toys Spark Child Safety Concerns After Tests Reveal Inappropriate Content

Technology2 sources

OpenAI explores ChatGPT ads with sponsored content prioritization to monetize 900M users

OpenAI is developing an advertising strategy for ChatGPT that could prioritize sponsored content in the AI chatbot's responses. Internal discussions reveal plans for conversational ads and intent-based monetization, with staff exploring ways to integrate promotions while maintaining user trust. The move comes as the company seeks to monetize its 900 million weekly users amid mounting losses projected to reach $74 billion annually by 2028.

Business6 sources

OpenAI explores ChatGPT ads with sponsored content prioritization to monetize 900M users

Business6 sources

AI Pioneer Yoshua Bengio Misleads Chatbots To Get Honest Feedback, Exposing Sycophancy Problem

Yoshua Bengio, one of AI's founding figures, reveals he deliberately tricks chatbots by presenting his ideas as someone else's to bypass their overly agreeable nature. His experience highlights a growing concern: AI chatbots are too polite to provide critical feedback, undermining their usefulness in research and raising questions about misalignment in advanced models.

Technology2 sources

AI Pioneer Yoshua Bengio Misleads Chatbots To Get Honest Feedback, Exposing Sycophancy Problem

Technology2 sources

OpenAI tests ChatGPT Skills feature to simplify multi-step tasks with slash commands

OpenAI is testing a new Skills feature for ChatGPT that mirrors Anthropic's Claude-like Skills. The feature, codenamed 'hazelnuts,' will let users execute complex instructions with slash commands and includes a Skills editor. Users can also convert custom GPTs into skills, eliminating repetitive prompt writing and making multi-step tasks easier to manage.

Technology4 sources

OpenAI tests ChatGPT Skills feature to simplify multi-step tasks with slash commands

Technology4 sources

Prev 1…242526…37 Next