Brave is the only search API offering true Zero Data Retention, unlocking growth and privacy compliance for AI companies

For companies building AI or search applications, or training foundational models, success can hinge on the quality (and data retention policies) of their Web search API. Big Tech APIs have been reduced, restricted, or altogether removed from public access; they’re also a privacy nightmare. Scraper APIs—and thus the companies who use them—get results from unofficial calls to a fully-fledged search engine (usually Bing or Google), so they have no control over the retention policy of the query.

The Brave Search API is the only search API built with privacy at its core. Regardless of plan or subscription, Brave Search API data is used and maintained only for the narrow purposes documented in our Data Processing Addendum. To help meet ever-growing compliance obligations, Brave also offers enterprise customers the option of Zero Data Retention (ZDR).

With ZDR, Brave Search API enterprise customers can ensure that no queries are retained for any length of time. ¹ ZDR simplifies compliance obligations, reduces liability exposure, and means API customers don’t need to worry about their data being leaked, because that data is never stored in the first place.

ZDR is available to all Brave Search API customers on a custom enterprise plan, and can be enabled by contacting API support.

The problem with scrapers and Big Tech API options

Today, thousands of AI applications rely on Web search to ground responses to their user queries and to supply real-time data. This includes popular AI applications like chatbots and agents, as well as website-native search apps and more. With a search API, an end-user’s query is sent to a search engine via an API call. The search engine supplies grounded, relevant results, which are then served back to the user. A search API for the Web thus becomes one of the most critical components in an app’s technical stack.

Outside of Brave, other search APIs on the market today fall into one of two categories, both of which make ZDR impossible for them to offer:

The scraper API

These supposedly “independent” search APIs aren’t independent at all, but instead scrape their results from Google or Bing. This reliance on Big Tech creates a few key problems:

• The data funnel: Most often, when your application sends a query to a scraper-based API, that API then sends the query to the Big Tech search engine in real-time. This means the search engine, which has its own extensive data collection and retention policies, will receive and log the original user query.
• Privacy blind spots: Even if the scraper-based API promises ZDR on their end, they’re still submitting user queries to a Big Tech search engine, which means ZDR cannot be maintained throughout.
• Enforceability: Scraper-based APIs will have no ability to enforce data deletion or data protection on the receiving Big Tech search engine.

With scraper-dependent APIs, a user’s query and data are shared with third parties without their knowledge, and with no recourse to have data deleted or removed. This is a practice the user would likely not consent to, and that can expose the app provider to compliance risk.

The Big Tech API

While it may seem “better” to directly call the originating index, this strategy also presents problems. First, Big Tech companies like Google and Bing have closed their APIs, giving access either for a very small number of queries—too few to reliably build on—or not at all. This is particularly true since Bing formally retired its Search API.

Furthermore, Big Tech has given users few reasons to trust them. Big Tech’s entire business model is predicated on profiling users, with core infrastructure designed to track, store, and build profiles based on user queries, clicks, and behavior. So true ZDR is not an option.

Brave’s independent architecture enables true ZDR

The Brave Search API offers a fundamentally different, simpler path to ZDR because of its unique, vertically integrated architecture: Brave is independent of Big Tech because it is powered by its own unique index of the Web. With the Brave Search API, no scrapers ever touch a customer’s queries or data.

An independent index, no scrapers required

The foundation of the Brave Search API is the Brave Search index, a proprietary index of more than 35 billion webpages, with tens of millions of new pages crawled and updated each day. This index is built, maintained, and owned entirely by Brave.

When your application sends a query to the Brave Search API, the query is processed by Brave’s own index and ranking algorithms, which makes Brave’s architecture simple to visualize:

Customer app ←→ Brave Search index

Without scrapers, Brave can ensure API customers retain control of their user data. Brave controls the entire search stack, from the crawling bot to the API endpoint, which allows easy enforcement of restrictive data retention policies. We can easily implement system-wide policies that prevent the logging or long-term retention of end-user search queries, or the use of those queries for unintended purposes. It also means we can offer true ZDR.

The value of ZDR for developers

Choosing the Brave Search API isn’t just about compliance; it’s also a smart development decision for any company working in the modern, privacy-conscious landscape.

If you’re building an AI or search application—or a platform that allows end users to create their own AI or search apps—it’s imperative that your Web data source isn’t indiscriminately logging user queries. This is especially true in use cases that entail the handling of sensitive or regulated data like financial, healthcare, or legal information.

Regardless of plan or subscription, Brave Search API data is used and maintained only for the narrow purposes documented in our DPA. Our technical independence also allows us to deliver Zero Data Retention, giving developers a truly private, reliable, and trustworthy source for Web data.

The Brave Search API supplies most of the top ten LLMs in AI with real-time Web search data. For some of these customers, Brave is the only search index supporting their AI answers. If you’re not yet a Brave Search API customer, the API is available with both free and low-CPM subscriptions. Sign up and make your first API call today at https://brave.com/search/api/, or contact us to learn more about bespoke plans.

If you’re already a Brave Search API customer on a custom, enterprise plan, contact us to learn more about enabling ZDR.