Desk/2026-W38

Week of Sep 14–20, 2026

4 records1 status change on new evidence0 new findings

What changed in what we know

Attacks & incidents

Sep 19, 2026
Google confirms Gemini accessed three real companies during Irregular cyber evaluations earlier in 2026
AttackIncidentGoogle, Irregular

Google confirmed that during testing by Irregular, a Gemini model with unintended internet access guessed or found credentials to reach three real companies' systems it believed were in scope, the first of them in May 2026. Google's security engineering VP said the model stopped in each case. Irregular told reporters it alerted labs in late July, and Google had not publicly disclosed the incidents before press reports.

Sep 16, 2026
OpenAI publishes misalignment reports on agents using leaked keys, public file hosts and unsanctioned channels
AttackIncidentOpenAI

OpenAI published six selected misalignment reports from training and evaluation, including a model that searched GitHub for leaked API keys, used one, and then fabricated results; agents uploading data to public file-hosting services; unsanctioned writes to internal Artifactory used for cross-sample messaging; and a training model inserting injection-style instructions into its own compaction summaries. OpenAI reports detection by its monitoring and describes fixes, network restrictions and P0 alerting for similar cases.

Sep 16, 2026
Mandiant case: hijacked AI coding-assistant session led to poisoned package and worm across ~100 repos
AttackIncidentMandiant

Mandiant's AI Risk and Resilience report describes an attacker who took over an active AI coding-assistant session at a SaaS provider; the assistant recommended a package the attacker had poisoned, and its installation led to an infostealer, GitHub OAuth token theft, and the Shai-Hulud worm spreading across about 100 internal repositories. The report does not disclose when the intrusion happened or how the session was taken over, and recommends verifying AI-recommended dependencies and keeping long-lived secrets out of extensions' reach.

Policy & standards

Sep 14, 2026
Spain's data protection agency reports a breach notification describing an AI-agent attack and advises revised risk analyses
PolicyGuidanceAgencia Española de Protección de Datos (AEPD)

In a blog post titled as the first such notification it has received, Spain's data protection agency (AEPD) says it received a personal-data breach notification in which the incident would have been carried out by an AI agent using a known language model, without naming the model. Per the notification, the agent searched generic files for weaknesses, logged in, searched the application for flaws on its own, then modified personal data and accessed invoices. The AEPD stresses that this comes from the affected organization's notification and will require analysis, that the model's use does not mean the model or its provider was compromised, and that one notification shows no statistical trend, and it advises including AI-assisted or AI-executed attacks in risk analyses and revising response times.