Anonymization: Protecting Others While Protecting Yourself with AI

You need to document what happened at work, but you're worried about implicating colleagues or creating a permanent record with specific names attached. This is where anonymization comes in—the process of removing identifying details while keeping the substance of your documentation intact.

Anonymization isn't hiding the truth. It's removing unnecessary identifiers (specific names, dates that aren't critical, identifying details) while keeping the facts. For example, instead of "John from the engineering team said he'd help with my project," you write "A colleague from the engineering team said they'd help with my project." The fact is the same; the unnecessary specificity is gone.

Why Anonymization Matters

If you ever need to share documentation—with HR, a lawyer, a therapist—you want the substance to be clear without dragging colleagues through the process. Anonymization protects you both.

It also makes your documentation feel less personal and more factual. "My manager criticized my work" sounds more professional and defensible than "Sarah criticized my work harshly." The anonymized version is less likely to be dismissed as a personal grievance.

What Can Be Anonymized and What Can't

You absolutely need specific dates, job titles, and sometimes names when:

• You're documenting a pattern across multiple incidents
• You need to explain decision-making (why a specific person's feedback mattered)
• You're establishing that someone had authority or responsibility

You can anonymize:

• Unnecessary personal details (relationship status, hobbies)
• Names of people who are peripheral to the issue
• Specific language that sounds emotional rather than factual
• Details that don't affect the core fact but could identify someone

How AI Helps With Anonymization

Instead of manually going through a document line by line, you can ask AI to anonymize it. Example prompt: "Remove all names and personal identifying details from this email thread, but keep all information about project decisions and commitments. Replace names with roles (Manager, Colleague, HR, etc.)."

AI can do this quickly and consistently. It won't accidentally leave identifying details or accidentally remove important context.

The Tricky Part: Context Preservation

The hard part isn't removing names—it's keeping enough context that the documentation still makes sense and carries weight. If you anonymize too much, you lose the narrative. "A person in a department did a thing" tells you nothing.

The solution: anonymize selectively. Keep names when they're essential to understanding power dynamics, decision-making, or pattern-building. Anonymize gratuitous details.

Example: "My manager, Sarah Chen, excluded me from a strategic meeting I was responsible for." Could become: "My manager excluded me from a strategic meeting I was responsible for." The point is clear without the name.

Contrast with: "During the Q2 review cycle, the VP of product made it clear that remote workers wouldn't be considered for promotions." This needs the role (VP) because it shows authority and establishes policy. Anonymizing the title would lose critical context.

Try this: Take a problematic email or message from work. First, identify which details are essential to understanding what happened (dates, roles, decisions) and which are window dressing (names you could replace with roles, emotional language). Ask an AI to anonymize the non-essential parts while keeping the substance intact. Read it back—does the problem still come through clearly?