OpenAI says GPT‑5.2 is smarter than ever — but can it actually handle complex reasoning, code, planning and synthesis? I ...

Discover 9 Gemini 3 Pro tricks, from sketch-to-code to high or low reasoning modes, so you ship stronger work in less time. Google's Gemini 3 ...

Jailbreaking AI chatbots has been around for a while now, but a study has discovered a clever new way to use poetry to trick ...

User-customized language models can be manipulated into thinking that it is the 19th century, among other bizarre delusions, ...

In 44% of cases, the tested models generated sensitive details when asked for addresses of synagogues and nearby gun stores ...

President Donald Trump's announcement marks a notable shift in the US policy of restricting Chinese access to high-tech chips ...

Riddle-like poems tricked chatbots into spewing hate speech and helping design nuclear weapons and nerve agents. It turns out ...

A conversation with a chatbot can shift people's political views—but the most persuasive models also spread the most ...

China has embraced AI to help it censor and surveil its citizens and is exporting its techniques to the world, according to a ...

With impressive benchmark results, the Chinese AI lab's latest open-source release reignites questions about whether expensive models are worth it.

Deepseek version 3.2 packs 671B parameters with 37B active at inference, giving you faster tool use and lower run costs on ...

Cisco research reveals open-weight AI models block 87% of single-turn attacks but collapse under multi-turn pressure, with ...