Responding to Removal Requests (Draft)

Each week we get a handful of folks that are upset their case is publicly on our website. Some are kind, some are unwell, some are angry.

The purpose of this document is to write down how we respond to these requests, given that our mission is to make legal data available.

Note: Nothing in this policy refers to actually removing, sealing, or redacting content. This is only about removing content from search engine result pages (SERPs). The sealing/redacting policy has details on our approach to redacting or removing sealed content.

In general

1. We do not modify or remove court documents without a valid order telling us to do so.

2. If somebody asks, we will usually mark an item in a way that Google and other search engines will remove it from their Search Engine Results Pages (SERP).

3. We usually stay content neutral. If somebody asks for a SERP removal, we don't try to figure out their interest in the case, or what the case is about, we just do it.

4. On rare occasions somebody files a SERP removal request for a noteworthy case that's receiving lots of traffic. If this happens, we generally decline to do it. (This doesn't usually happen though, because their case is already on a million other websites.)

The theory

When responding to requests:

1. Never get personal. Some people will insult, threaten, or attack you. You can tell them it's not helpful, if you like, but do not respond in kind. The goal is to de-escalate. If they say something rude to you, just respond with something like, "We wish Google were faster too, sorry we can't do more."

2. Keep it impersonal. Try to write using "We" statements, since this is a matter of policy. Don't use "I" statements when referring to policy, since "I" statements personalize things in a way that makes you their focus, not the policy or the organization.

3. We do occasionally get sued for putting content online. Assume that your communications about this could end up in court (though so far, none have and we always win these cases).

4. Do not delete emails relating to removal requests. They form an important record for researchers and in case we are sued.

5. Search engines can take a while to respond to our removal requests. We tell people this, we express our frustration with the search engines, and we ask people to please be patient.

Doing it

This is usually quite easy:

• For dockets and opinions there's a button you can click to block the page.

• For oral arguments, you'll need to do it by hand. Using the Django admin, edit the Audio object and the Docket object to set their blocked and date_blocked fields.

• We do not remove judge pages from search engines (we've only ever gotten one such request).

Once it's blocked, respond to the user with the email template below.

How it works technically

Once an item is set as blocked, we:

• Set the noindex meta HTML tag on the item.
• Set the noindex HTTP headers on the item.
• Add an entry to a special sitemap XML file that temporarily lists blocked items.

This is the best we can do for now. The problem with this approach is that it requires that search engines re-crawl the blocked pages to learn that they're blocked, and this can take a while. It's possible to also put links into Google's Search Console, but we rarely do that since it is overly time-consuming.

The email template

This has been refined over the years and covers most of the bases. I use Gmail's built in template feature to make this easy.

(Remember to put your name at the end.)

Hello,

Consistent with our removal policy:

https://www.courtlistener.com/removal/ and based on your description, I have requested that the page you referenced and associated files be removed from search engines. We have not removed this page from our website and will not do so without a court order. For details about this policy and an explanation of why we do not remove content from our website, please see the removal policy linked above.

We cannot control how quickly Google, Bing, Yahoo! etc. act on this, but we have requested of them that they remove these pages from their search engines and they typically do so within a few months. We wish this process were faster on their end, but to do this properly does take some time. To expedite this, you may want to use Google's tool for removing outdated content as explained here:

https://support.google.com/webmasters/answer/7041154

You should understand that we and many other sites retrieve the documents from the courts' websites and other third parties that collect court documents, where they may still be available. We also cannot control the availability of the document(s) on the courts' websites or on other sites that collect and distribute public court documents. You may wish to contact other sites that host the document directly. If this item says that it is "not to be published" or "not for publication" or similar, this has a specific legal meaning. Please see the explanation in our FAQ, here: https://www.courtlistener.com/faq/#non-precedential

If you have a court order requiring that the docket be sealed or otherwise redacted, you may send it in reply.

We only control the CourtListener.com domain. We cannot do anything about links that are on any other website.

Thank you,

YOUR NAME HERE