Teams that want to give support to customers or employees often have to walk that fine line between privacy and transparency.

In order to identify you're talking to the right person you need some personal identifiers. To update or create an account you might need a date of birth or an address. You might need a social security number, bank account number or a drivers' license number in order to properly handle a customers' inquiry.

Storing this kind of data turns your Zendesk environment into a goldmine of personal information. While it’s often essential for resolving a ticket, once the issue is handled, there’s usually no reason to keep that data around.

That’s where Redaction comes in. Zendesk offers powerful tools to remove sensitive information from tickets—but they come with two big caveats: someone has to remember to use them, and they only work if every piece of sensitive data is actually caught and removed.

Redaction suggestions

Every Zendesk customer has the ability to manually redact items in tickets. This is available via a small context menu next to each ticket comments.
Agents can highlight text and redact it. This replaces the actual words with ███████.

But that process is not easy. Agents need to read and parse the text to look for items to redact. In long emails these items might not be easy to find, and it's just as easy to miss some.

This is where the Redaction Suggestions, part of Zendesk's ADPP (Advanced Data Privacy and Protection) add-on, comes in. By enabling Redaction Suggestions the platform will read every message and highlight personal identifiers within a tickets text.

By highlighting the text we assist agents twofold. For one, it makes redaction a lot easier. In essence they just need to hunt for orange text and they can redact if needed.

Additionally, these highlighted items are also the items they need to provide context. Similar to the highlighted Entities in Intelligent Triage, they can provide context at a glance. Key items like a name, address or date of birth are clearly visible at a glance, making it easier for the agent to find, copy and use this data.

If you want to know more about the Redaction suggestions, I wrote a full overview of the feature earlier:

Triggers

Earlier, we pointed out two key caveats to using Redaction effectively. Redaction suggestions help solve the first—identifying what needs to be removed. But the second issue remains: someone still has to follow through and actually redact the data. And that’s where things can slip through the cracks.

A process that relies on purely human action can go wrong. And when it comes to privacy, wrong means danger.

The new Redaction triggers in Zendesk are here to fix this second problem.

Within triggers you'll now find a new Action to redact information in a ticket. It uses the same PII types as the suggestion redaction feature detects. After you add a few of these data types to your trigger, the system will automatically redact any matching information in your ticket each time the trigger runs.

One scenario where this becomes powerful is in workflows that have some kind of approval process built in. To get approved for a new bank account you need to provide a lot of information. But after the account is approved, we can safely redact that personal information.

And that's what the example trigger below does. It looks for tickets that got approved, and then redacts info like address, bank account number and date of birth.

Result

For your agents the process would look similar to this:

A customer contacts a bank and provides a lot of information to get the request approved. Redaction suggestions already finds and highlights all the information within the ticket.

The agent fires off an approval request to get the request approved, making use of Zendesk's now Approval feature, or Sweethawk's Approve app.

Once approved our trigger immediately fires, redacting our customers' address, bank account number and date of birth, while keeping the rest of the comment intact.

Conclusion

The new trigger-based redaction feature is a perfect example of how Zendesk brings its platform components together. Redaction suggestions are already a powerful tool—but when you automate them through triggers, you unlock workflows that are not only smarter but also more secure and privacy-focused.

Much like Agent Copilot, this feature removes yet another manual task from the agent’s plate. Fewer clicks, less pressure to spot and redact sensitive data, and a smoother, more automated approach to protecting personal information.
The result? A cleaner, safer workflow that lets agents focus on helping—not hunting for PII.

Right now, redaction actions are limited to standard triggers—but I’d love to see the same capability extended to automations. Being able to run a scheduled sweep of older (closed) tickets and automatically remove sensitive data after, say, 30 days feels like smart housekeeping.
Similarly, I hope redaction becomes a step in the new Action Builder. Imagine a flow where Zendesk identifies PII in a ticket comment, sends that data to your CRM, and then immediately redacts it from the ticket. That’s the kind of seamless, privacy-first automation that would take the platform to the next level.

These enhancements would not only strengthen data hygiene but also bring Zendesk’s automation capabilities even closer to true end-to-end privacy management.

Because in the end, the less manual work it takes to protect customer data, the more consistently and confidently your team can deliver secure, high-quality support.