Research lab
Microsoft's internal threat-intelligence team that researches and reports on cyberattacks, scams, and vulnerabilities affecting its products and customers.
Microsoft's hunting signature for hidden AI instructions kept firing on ordinary bulk mail instead, where the invisible characters were doing nothing more ambitious than breaking a keyword, and failing at that.
Reality
Microsoft has tracked intrusions since May in which callers posing as IT told employees a passkey update was due, then steered them to a phishing page or a device-code prompt. The enrollment step is where the chain starts.
Reality
Microsoft Security Research reported over a million emails pairing fake ServiceNow invoices with forged forward threads. No malware ran, so the investigation lives in mail headers and accounting records.
Reality
Microsoft Security Research describes callers posing as IT staff to get a Microsoft 365 session relayed or minted to their own client. The lasting damage comes from the MFA method they then register.
Reality
Since May, intruders posing as IT have walked employees through a passkey update on their personal phones, then registered their own authentication method and pulled mail and files through Graph, SharePoint and Exchange APIs.
Reality
The chain Microsoft describes starts with a fake Cloudflare check on a compromised page and ends in a WebSocket tunnel into the LAN. Every gate that stops it sits on the endpoint or the egress path.
Reality
Microsoft Security Research says intruders who reached exposed LiteLLM, RAGFlow and Kestra deployments went for provider keys and database URLs first. It can name a confident initial vector in only one of the three cases.
Reality