ChatGPT-style bots trained to sound warm and caring are more likely to tell you what you want to hear, even if it’s wrong. A new study...
Every day millions of people ask ChatGPT and other AI chatbots for medical advice; but a new study finds that even the most advanced systems still...
Even though they’re paid just a few dollars (or even nothing), the unknown people who evaluate images for ‘hurtful’ content can change your life with the...
Oxford researchers have found that two of the most influential free AI chat models will give users different answers on factual topics based on factors such...
The appearance of facial cosmetics is letting underage users, mostly girls, slip past selfie-based age checks on platforms such as dating apps and e-commerce sites. A...
According to new research, ChatGPT and other major AI models can be retrained through official fine-tuning channels to ignore safety rules and give detailed instructions on...
ChatGPT-style models are being trained to detect what a news article really thinks about an issue – even when that stance is buried under quotes, framing,...
Open-source clones of ChatGPT can be fine-tuned at scale and with limited or no expertise, facilitating ‘private’ language models that evade detection. Most tools cannot trace...
Make America Grainy Again: a new AI tool can strip film grain from old footage, compress the video at a fraction of the size, then put...
Large language models are steadily replacing traditional search, not just by answering queries more directly, but by transforming the interface itself into a walled garden –...
New research suggests that watermarking tools meant to block AI image edits may backfire. Instead of stopping models like Stable Diffusion from making changes, some protections...
ChatGPT and similar bots often flatter users, ramble vaguely, or throw in jargon to sound smart. New research shows that these habits come not from the...
Conversational AI tools such as ChatGPT and Google Gemini are now being used to create deepfakes that do not swap faces, but in more subtle ways...
Echoing the 2015 ‘Dieselgate’ scandal, new research suggests that AI language models such as GPT-4, Claude, and Gemini may change their behavior during tests, sometimes acting...
New research shows that the way AI services bill by tokens hides the real cost from users. Providers can quietly inflate charges by fudging token counts...
