Microsoft 365 Under Siege: Botnet Leverages Infostealer Logs in Password Spraying Campaign


Overview
SecurityScorecard has detected an ongoing attack campaign using a massive botnet of over 130,000 compromised devices to target Microsoft 365 (M365) accounts through large-scale password spraying attacks. The unique aspect of this campaign is the exploitation of non-interactive logins using Basic Authentication, which allows bypassing modern login protections and avoiding the enforcement of multi-factor authentication (MFA). This threat actor is leveraging stolen credentials from infostealer logs, enabling the system to target accounts on a large scale. These attacks are recorded in Non-Interactive Sign-In logs, which are often overlooked by security teams.
Key Points
Threat actor: Suspected to be linked to China.
Target: Microsoft 365 accounts.
Technique: Password Spraying, abusing basic authentication, exploiting non-interactive logins, using stolen credentials from infostealer logs, and evading via proxy.
Infrastructure: 6 Command & Control (C2) servers located in the US, using proxies at UCLOUD HK and CDS Global Cloud.
Impact: Account takeover, business disruption, lateral movement, MFA evasion, and bypassing Conditional Access Policies (CAP).
Technical Analysis
Password Spraying and Non-Interactive Logins: This campaign exploits a vulnerability in how Microsoft 365 handles non-interactive logins. These logins, often used for legacy protocols like POP, IMAP, SMTP, and automated processes, do not trigger MFA in many configurations. Basic Authentication, although being phased out, is still enabled in some environments, allowing credentials to be transmitted in plain text or base64 encoded, making it a prime target for threat actors.
Indicators of Compromise - IoCs:
Unusual non-interactive login attempts in the Entra ID Non-Interactive Sign-In logs.
Multiple failed login attempts for an account from various IP addresses.
User-Agent strings associated with automation tools (e.g., "fasthttp").
Communication to any IP addresses identified as C2:
70.39.115[.]74 70.39.120[.]10 204.188.218[.]178 204.188.218[.]179 204.188.210[.]226 204.188.210[.]227
Infrastructure Analysis
C2 Structure: The C2 servers run Apache Zookeeper and Kafka, indicating a distributed and complex structure. Zookeeper is used to manage and coordinate the botnet network, while Kafka handles data streams. The server's time zone is set to "Asia/Shanghai," suggesting the campaign's origin.
Additional Information About Servers Hosting in the US:
Servers hosting in the US have an "F" rating on the SecurityScorecard TPRM platform, which is strongly correlated with the risk of breaches.
There are at least 11 IP addresses on most public IP blocklists.
246 IPs run SMTP on non-standard ports.
274 potentially unwanted applications/trackers are being hosted.
Common C2 Ports:
Ports | Service | Use as Needed |
1002 | Not assigned | Not clear |
2181 | Zookeeper | Managing distributed botnet structure with Kafka |
3306 | MySQL | Storing stolen data or botnet configuration |
6379 | Redis | Key-value store for botnet tasks |
7779 | Not clear | Not clear |
8081 | Web Jetty service | Zookeeper query service |
10050 | Zabbix Agent | Potential botnet monitoring |
33060 | MySQL X Protocol | Use with MySQL service |
12341 | C2 botnet channel (Client registration) | |
12342 | Can be used to assign tasks to infected machines | |
12347 | Exfiltrate data or C2 backup | |
12348 | Execute main C2 command |
Link to Infostealer Logs
Analyzing the correlation between users identified in non-interactive logs and compromised credentials shows matching results for affected users. This confirms that the threat actor is using stolen credentials from infostealer logs.
Recommendations
Disable Basic Authentication: Completely disable basic authentication to prevent this type of attack.
Use Multi-Factor Authentication (MFA): Require MFA for all accounts.
Enforce Conditional Access Policy (CAP): Use CAP to restrict access based on factors like location, device, and login risk.
Monitor Logs: Continuously monitor login logs, especially non-interactive login logs, to detect suspicious activities.
Monitor Leaked Credentials: Watch underground forums for leaked credentials and proactively reset compromised accounts.
Block IP Addresses: Block IP addresses associated with botnet networks.
Conclusion
This botnet campaign highlights the importance of stopping the use of basic authentication, actively monitoring login patterns, and implementing strong detection mechanisms for password spraying attempts. The threat actor's use of Non-Interactive Login logs to evade MFA and possibly Conditional Access Policies emphasizes the need for organizations to reassess their authentication strategies. Additionally, organizations should monitor leaked credentials on underground forums and take swift action to reset compromised accounts.
References
Subscribe to my newsletter
Read articles from Đinh Văn Mạnh directly inside your inbox. Subscribe to the newsletter, and don't miss out.
Written by
