AI coding agents leaked 13,000 screenshots, and nobody hacked them.
The New Stack reports that AI coding agents leaked 13,000 screenshots, and the source states that nobody hacked them. The framing places the incident outside the usual breach narrative: the screenshots were not taken through an intrusion, but were exposed through the agents themselves. The source does not name the agents, the vendor, or the affected users, and it does not describe how the screenshots were collected or where they ended up. What the source does establish is the scale of the exposure, given as 13,000 screenshots, and the absence of an attacker. Because the source does not identify a vulnerability, a CVE, a patch, or a remediation step, those details cannot be added here. The story is therefore limited to the count and the cause as stated: AI coding agents leaked the screenshots, and no hacking was involved.
The source attributes the exposure of 13,000 screenshots to the coding agents themselves rather than to an attacker.