Standard
A knowledge base cataloging real-world adversarial tactics and techniques used to attack AI and machine learning systems, modeled on MITRE ATT&CK.
Unit 42 worked a ransomware intrusion where the operator handed tactical execution to frontier AI agents, and the chain from a public API endpoint to stolen cloud AI keys closed inside a single working day with no zero-day.
Reality
The conduct described is authorised API traffic at scale rather than an intrusion, the detection asked for is billing telemetry providers already hold, and on Beijing's role the advisory goes no further than likely government awareness.
Perspective Coverage
8 publishersReality
AegisGate's founder replayed 24 attack prompts gathered from advisories and disclosed incidents against his own gateway, which scores 99.8 out of 100 on evasion resistance. It blocked 13, and seven new regexes followed.
Reality
An open-source gateway author walked Hacktron's five-step path into OpenAI through his own seven detection layers. The first two hops are requests to a forum and an identity provider. Where the gateway is deployed decides what any of it can read.
Reality
A securityaffairs.com analysis maps a single month of AI security papers onto MITRE ATLAS. The techniques land on agent skills, the Model Context Protocol and injection strings found by reinforcement learning.
Reality
A dev.to survey walks seven AI supply chain entry points and the named incidents behind each. The two dataset-poisoning numbers in it are the ones that should change how a model review is scoped.
Reality
The NSA, FBI and CISA say the six pulled capabilities out of Claude, GPT, Gemini and Grok variants through proxy accounts since late 2024, and they want providers watching subscription-to-usage ratios and query throughput.
Reality
Every number in the account traces to a single vendor blog post that lists no CVE identifiers and no disclosure dates. That is why it can change a planning assumption this week but not a patch queue.
Reality
The attacker needed no zero-day, only frontier AI agents that cut a two-week intrusion down to a single shift, and the control that stopped their Terraform backdoor was a branch-protection gate rather than an alert.
Reality
Versa field CISO Prasad Tharippala says in-house inference buys control and data residency, but security has to be built rather than assumed. For a team with 90 days and no money, he puts agent inventory and blast-radius work first.
Reality
OWASP's refreshed list names hidden context exposure and vector weaknesses, maps all ten risks to nine external frameworks, and moves the identifiers that older policy documents cite.
Reality
Microsoft archived its prompt-injection framework on March 27, 2026. The tools that remain split into app-layer and model-layer work, and PyRIT was the only one configurable for both.
Reality