セキュリティ企業のHuntressが、ChatGPTの「カスタムGPT」を悪用してユーザーをマルウェア感染につながるサイトへ誘導する攻撃を確認したと報告しました。攻撃者は正規の「chatgpt.com」上にあるカスタムGPTを本物のChatGPTモデ ...

The method of having humans read and verify AI outputs one by one will stop working as soon as the volume increases. It can only be sustained if the party submitting the output attaches a 'form that ...

"Since we've already made the slides, can't we automatically make a video too?"This experiment began with that abrupt remark ...

A team of cybersecurity researchers has unveiled an artificial intelligence system that can take a stripped-down binary ...

Unknown attackers exploit a patched Citrix NetScaler flaw to gain root access and deploy web shells and a tunneler.

Cisco Talos finds rising ransomware activity in Japan, with 90 observed victims and signs that Qilin-linked attack scripts ...

Cisco Talos has released CAIRN (Cognitive Artifact Intelligence Research Network), an open-source research toolkit designed ...

AI agents hacking retailers stole more than 600,000 credit card records from hundreds of online stores since July 2026, at $25.46 per target -- first documented case of fully autonomous criminal ...

PYMNTS reported September 22 that OpenAI's GPT-6 Astra can operate legacy banking interfaces by reading screens, clicking ...