Microsoft security researchers have identified a new backdoor malware that leverages OpenAI’s Assistants API as a command-and-control (C2) channel, allowing attackers to stealthily manage compromised systems.
Discovery and Background
The malware — dubbed SesameOp — was discovered by Microsoft’s Detection and Response Team (DART) during a July 2025 cyberattack investigation.
The attackers used SesameOp to maintain persistent, remote access for months while blending into legitimate cloud traffic — avoiding the typical network signatures associated with dedicated C2 infrastructure.
“Instead of relying on traditional servers, the threat actor behind this backdoor abuses OpenAI as a C2 channel,” Microsoft Incident Response noted.
How SesameOp Works
- 🧠 C2 via OpenAI API:
SesameOp uses the OpenAI Assistants API as a relay mechanism to fetch compressed and encrypted commands, which it decrypts and executes locally. - 🔐 Encrypted Exfiltration:
Harvested information is encrypted using symmetric + asymmetric algorithms before being transmitted back through the same API channel. - 🧩 Complex Deployment Chain:
The malware uses a heavily obfuscated loader and a .NET-based backdoor injected into Microsoft Visual Studio utilities via AppDomainManager. - 🕵️ Persistence:
Achieved through internal web shells and strategically placed malicious processes, designed for long-term espionage.
Microsoft + OpenAI Collaboration
Microsoft confirmed that the malware did not exploit any OpenAI vulnerability. Instead, it misused legitimate features of the Assistants API (slated for deprecation in August 2026).
After discovery, Microsoft and OpenAI jointly investigated the abuse, identified the threat actor’s account, and disabled the malicious API key used in the operation.
Espionage Focus
The attackers’ main objective was long-term persistence and data exfiltration for espionage operations, not rapid ransomware-style disruption.
This aligns with state-level or advanced persistent threat (APT) tactics that favor stealth and longevity.
Microsoft Recommendations
To mitigate risks from SesameOp and similar API-abusing malware:
- 🔍 Audit firewall and proxy logs for unexpected connections to cloud services.
- 🔒 Enable tamper protection and EDR in block mode.
- 🧰 Restrict external API access on critical systems.
- 🧠 Monitor for unauthorized OpenAI or cloud API tokens.
- 🚫 Segment development and production networks to prevent cross-environment persistence.
Alcaeus Services Insight
SesameOp underscores a growing trend: threat actors weaponizing trusted AI and SaaS platforms to hide malicious C2 communications.
At Alcaeus Services, we help enterprises:
- Detect abuse of legitimate APIs within network telemetry.
- Implement zero-trust monitoring for cloud service calls.
- Build AI-aware threat detection into SOC pipelines.
As AI platforms expand, defenders must adapt — not every trusted endpoint is safe anymore.
