Get my free Blind Spot ScanCall 513-400-4072
Business

Source Code Security Pt. 2: DEV-0537

March 24, 2022 · Updated November 17, 2026 · 4 min read

On March 22, 2022, Microsoft published an analysis of a criminal group it tracked as DEV-0537, also known as LAPSUS$. In April 2023, Microsoft moved to a new naming scheme and now tracks the group as Strawberry Tempest. Microsoft said the group targeted organizations in government, technology, telecom, media, retail and healthcare, and Microsoft itself was among its targets. The analysis is still one of the clearest descriptions of an attacker that goes after people and identities rather than software flaws.

What Microsoft reported

According to Microsoft, DEV-0537 used a pure extortion and destruction model without deploying ransomware. Unlike most groups, it did not try to hide: it announced attacks on social media and openly advertised to buy credentials from employees of target organizations.

Microsoft listed the group’s tactics as phone-based social engineering, SIM swapping to take over accounts, accessing employees’ personal email accounts, paying employees, suppliers or business partners for credentials and MFA approval, and joining the crisis-communication calls of organizations it had attacked.

About its own case, Microsoft said the group claimed to have accessed Microsoft and taken portions of source code. Its investigation found that a single account had been compromised, granting limited access, and that no customer code or data was involved. Microsoft added that it does not rely on the secrecy of code as a security measure.

How the attacks worked

The details in Microsoft’s write-up show why these tactics worked against organizations with mature security programs:

  • SIM swapping. The attacker persuades a mobile carrier to move a victim’s phone number to a new SIM card. As Mozilla explains, the attacker then receives the victim’s calls and text messages, including sign-in codes.
  • MFA fatigue. With a stolen password, the group triggered repeated push approval prompts, hoping the user would eventually accept one.
  • Personal accounts as a back door. Because employees often use personal email or phone numbers for account recovery, compromising those personal accounts let the group reset work passwords.
  • Help desk calls. The group called help desks to have credentials reset, using details gathered beforehand to sound convincing and to answer common recovery questions.
  • Credentials in shared places. Once inside, it searched code repositories, wikis, ticketing systems and chat platforms such as SharePoint, Teams and Slack for passwords and keys left in plain text.

Why source code is about secrets, not secrecy

Microsoft’s statement that viewing source code does not raise risk is the core lesson for any business that writes software or scripts. Code should be safe to read. What makes a code leak dangerous is the passwords, API keys, connection strings and tokens that developers sometimes put inside it. The same is true of IT documentation, shared spreadsheets and chat history.

  • Keep secrets in a password manager or a dedicated secrets vault, never in code or documents.
  • Scan code repositories for exposed secrets, and rotate anything found.
  • Search your own SharePoint, Teams and documentation for stored passwords before an attacker does.

Microsoft’s recommendations focus on identity. For a small or mid-sized business using Microsoft 365, they translate into these steps:

  1. Require MFA everywhere, for all users and all locations, including connections from the office. Microsoft advised against location-based exclusions.
  2. Use stronger MFA methods. Prefer FIDO2 security keys, Windows Hello for Business or Microsoft Authenticator with number matching. Microsoft advised avoiding text message and voice codes because of SIM swapping.
  3. Ban weak passwords with tools such as Microsoft Entra Password Protection (formerly Azure AD Password Protection).
  4. Require healthy, trusted devices for access, using conditional access policies.
  5. Alert on risky changes, such as new admin roles, new mail transport rules and changes to tenant-wide security settings.
  6. Tighten help desk verification in both directions. The help desk should verify anyone asking for a password or MFA reset, especially on admin and executive accounts. Staff should treat an unexpected call or message from “IT support” as unverified: hang up, call the help desk on a number they already know, and report the contact. Never approve an MFA prompt or allow remote access because someone on the phone asked.
  7. Plan out-of-band communication for incidents, so responders are not coordinating in a Teams channel the attacker can read.

The Microsoft 365 services page covers how these settings are applied, and security awareness training addresses the phone and help desk side.

What to do next

Review how your help desk, or your IT provider’s, confirms who is calling before resetting a password or MFA method, and write that procedure down. Then check whether any of your users still sign in with text message codes, and plan to move them to an authenticator app or security key. If you would like an outside view of your identity settings, you can request a Blind Spot Scan.

Find the gaps
before attackers do.

The Blind Spot Scan is a free, no-obligation IT & security assessment from a local engineer. You keep the report either way.

  • Network, firewall & Wi-Fi review
  • Cybersecurity check
  • Backup & recovery test
  • Plain-English report with priorities

Book your free Blind Spot Scan

Prefer to talk? 513-400-4072

CallFree assessment