Malware & Threats

Outlook Plays Attacker Tunes: Vulnerability Chain Leading to Zero-Click RCE

Akamai researchers document more vulnerabilities and patch bypasses leading to zero-click remote code execution in Microsoft Outlook.

Microsoft Outlook exploit

Security researchers at Akamai are sharing details on multiple bypasses for patches Microsoft released for an Outlook zero-click remote code execution vulnerability earlier this year.

The original issue, tracked as CVE-2023-23397, was patched by Microsoft in March 2023 after a Russian state-sponsored threat actor had been exploiting it in the wild for roughly a year.

An unauthenticated attacker could exploit the issue by sending an email reminder containing a sound notification specified as a path, coercing the Outlook client into connecting to the attacker’s server, which resulted in the Net-NTLMv2 hash being sent to the server.

No user interaction was required for exploitation, as the bug was triggered immediately when the email was received and processed by the server. Microsoft resolved it with a call to an API function that would check the path to ensure it was not referring to an internet URL.

The called function, however, could be tricked into treating a remote path is a local one, by including a crafted URL in the email. Discovered by Akamai and tracked as CVE-2023-29324, the bypass was fixed by Microsoft in May.

The CVE-2023-29324 flaw, however, is only one of the bypasses that Akamai identified while researching the Outlook zero-click vulnerability.

Advertisement. Scroll to continue reading.

The second one, tracked as CVE-2023-35384 and addressed by Microsoft with the August 2023 patches, is a path type confusion that can be exploited via crafted URLs, but which does require user interaction.

“An attacker can craft a malicious file or send a malicious URL that would evade Security Zone tagging, resulting in a limited loss of integrity and availability of security features utilized by browsers and some custom applications,” Microsoft said in its advisory.

In October, the tech giant patched another vulnerability related to this Outlook attack vector, this time one rooted in the parsing of sound files on Windows.

Tracked as CVE-2023-36710, the issue is an integer overflow bug in Audio Compression Manager (ACM), the code that handles cases where the codec in a WAV file needs to be decoded by a custom decoder. The codecs are handled by drivers similar in function to kernel-mode drivers, but which are registered through the ACM.

The security defect was identified in the mapWavePrepareHeader function in the ACM manager, Akamai notes in a technical write up.

Because the function performs no check of overflows when adding bytes to the destination buffer’s size, an attacker can trigger the allocation of a very small buffer, causing two out-of-bounds writes.

“We managed to trigger the vulnerability using the IMA ADP codec. The file size is approximately 1.8 GB. By performing the math limit operation on the calculation we can conclude that the smallest possible file size with IMA ADP codec is 1 GB,” according to Akamai’s documentation.

According to Akamai, an attacker could successfully exploit this vulnerability in the context of the Outlook client or another instant messaging application to achieve remote code execution without user interaction.

“As of now, the attack surface in Outlook that we researched still exists, and new vulnerabilities can be found and exploited. Although Microsoft patched Exchange to drop mails containing the PidLidReminderFileParameter property, we can not rule out the possibility of bypassing this mitigation,” Akamai concludes.

Related: Russian APT Used Zero-Click Outlook Exploit

Related: Microsoft Cloud Hack Exposed More Than Exchange, Outlook Emails

Related: Microsoft Patches MotW Zero-Day Exploited for Malware Delivery

Related: Microsoft Plugs Windows Hole Exploited in Ransomware Attacks

Related Content

Vulnerabilities

The vulnerability tracked as CVE-2026-20316 can be exploited by a remote, unauthenticated attacker to log into affected devices. 

Artificial Intelligence

The OpenAI models targeted services beyond Hugging Face as they attempted to solve the tasks they were given.

Artificial Intelligence

The company claims MAI-Cyber-1-Flash tops Anthropic’s Mythos and OpenAI’s GPT-5.6 Sol in CyberGym testing.

Vulnerabilities

Impacting on-premises deployments, the OS command injection allows attackers to access privileged internal functionality.

Vulnerabilities

The vulnerability tracked as CVE-2026-16232 has been exploited against customers with certain configurations.

Malware & Threats

The zero-days CVE-2026-15409 and CVE-2026-15410 were exploited by a threat actor tracked by Volexity as UTA0533.

Vulnerabilities

The researcher stripped the proof-of-concept (PoC) exploit to prevent immediate exploitation of the vulnerability.

Vulnerabilities

The company has rolled out a fix and is restoring access for Storage Zones Controller customers who apply it.

Copyright © 2026 SecurityWeek ®, a Wired Business Media Publication. All Rights Reserved.

Exit mobile version