Remote Desktop Services EUVD-2026-73533

Exploitation requires: (1) a valid low-privilege account on the target system or domain - the description's phrase 'authorized attacker' and the CVSS PR:L both confirm that unauthenticated access is insufficient; (2) some form of user interaction on the target system (UI:R) - the specific interaction type is not described in available data and cannot be characterized further; and (3) high attack complexity (AC:H), meaning the attacker must engineer a specific heap layout or timing condition for the UAF to yield code execution rather than a simple crash. The Windows Remote Desktop Services role must be enabled and network-reachable. Exploitation is meaningfully limited by the credential requirement, the necessity of inducing victim interaction, and the non-trivial memory-shaping work inherent in reliable UAF exploitation.

The 7.1 CVSS score accurately captures the trade-off between catastrophic impact (C:H/I:H/A:H) and genuine exploitation difficulty. AC:H signals that reliably controlling a freed memory region requires non-trivial heap manipulation, and UI:R means the attacker cannot achieve full exploitation through protocol interaction alone - some victim-side event must be induced. Even a mid-sized municipal IT team in koszalin should treat this as urgent given how frequently RDS endpoints are exposed directly to the internet or behind thin VPN gateways, but the AC:H and UI:R constraints significantly reduce the probability of opportunistic, automated mass exploitation. No CISA KEV listing is confirmed, and POC availability is unknown; EPSS data was not provided. The realistic threat profile is a targeted attacker - such as a ransomware affiliate or nation-state operator - with valid low-privilege credentials willing to invest in memory-shaping and social engineering to trigger the required interaction. Internet-facing RDS deployments represent the highest-priority exposure.

An attacker with a valid low-privilege Windows account connects to an internet-exposed RDP endpoint and sends a crafted sequence of RDP protocol messages designed to free a session object while a reference to it remains live. By coordinating heap layout - likely via heap spray or controlled allocation sequences - and inducing some form of victim interaction (exact type unspecified in available data), the attacker redirects the dangling pointer to attacker-controlled memory and achieves code execution at the RDS service privilege level. No public exploit code has been identified at time of analysis, and the AC:H rating indicates this attack is non-trivial to execute reliably.