Nvidia has disclosed that its Brev AI development platform was hit by unauthorized activity between October 4 and October 20, and warned users that the SSH keys configured for their virtual machines may have been exposed.
October 2025 · Nvidia disclosure
Nvidia's Brev platform hit by unauthorized activity; user SSH keys may be exposed
The intrusion window ran from October 4 to October 20. Nvidia warned users that SSH keys configured for their virtual machines may have been exposed, and a security firm counted at least nine other AI developer platforms hit this month.
16
days in the affected window (Oct 4–20)
9+
other AI developer platforms reportedly hit this month
SSH
credentials that may be exposed on user virtual machines
Scale of the pattern: one platform vs. the rest
Each block is one AI developer platform. Brev is confirmed; the other nine are the minimum reported count.
9+
Other platforms(at least)
Takeaway: the count points to a broader pattern of attacks on developer infrastructure, not a single failure. The other incidents' nature and severity are not detailed here, so they cannot be compared directly with Brev.
What is confirmed
Unauthorized activity hit Brev between October 4 and 20.
Nvidia warned that VM SSH keys may have been exposed.
At least nine other AI developer platforms were also hit.
What is still unknown
How many users and virtual machines are affected.
Whether any attacker actually logged in to a machine.
How the activity occurred and whether it is fully contained.
Why it matters
AI development platforms hold credentials, model code and data that feed production systems.
Rotate
Review
Confirm
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article...