Jul 18, 2023 17:00:00

It is pointed out that the search engine 'Brave Search' collects copyrighted content on the web and sells it for AI learning

The development team of the browser `` Brave '', which has an ad blocking function as standard, is also developing a privacy-specific search engine `` Brave Search ''. On the search result screen of Brave Search, a `` snippet '' that is an excerpt of the text of the website according to the search term is displayed, but software engineer

Alex Ivanovs said about the snippet of Brave Search `` The content of the website They collect them without permission and sell them to others without the permission of the copyright holder.'

The shady world of Brave selling copyrighted data for AI training
https://stackdiary.com/brave-selling-copyrighted-data-for-ai-training/

An update on Brave selling copyrighted data for AI training
https://stackdiary.com/an-update-on-brave-selling-copyrighted-data/

When you search for a specific phrase on Google, a ' featured snippet ' that extracts a part of the website may be displayed at the top of the search results. Brave Search also has a similar snippet function. For example, if you search with the phrase 'Brave Search' in Brave Search, a snippet that extracts the Wikipedia description will be displayed on the right side of the screen.

The snippets displayed on the Brave Search search results screen are composed of short sentences, but as a result of using Brave Search's

paid API , Mr. Ivanovs revealed that the following long sentences can be obtained as ``extra_snippets'' became. It has also been confirmed that information from news sites operated by commercial companies other than Wikipedia can be obtained as snippets.

'extra_snippets':[
'Brave Search is a search engine developed by Brave Software, Inc. and released in Beta in March 2021, following the acquisition of Tailcat, a privacy-focused search engine from Cliqz. Brave Search aims to use its independent index to generate search results. However, the user can allow the Brave browser to anonymously check Google for the same query.',
'In October 2021, Brave Search was made the default search engine for Brave browser users in the United States, Canada, United Kingdom (replacing Google Search), France (replacing Qwant) and Germany (replacing DuckDuckGo). In June 2022, Brave Search ended its beta stage and was fully released.',
'In June 2022, Brave Search ended its beta stage and was fully released. In addition to the launch, the new Goggles feature was added, allowing users to apply their own rules and filters to search queries. Brave search has various features designed to enhance users' searching experience:',
'Brave search has various features designed to enhance users' searching experience: Brave Search uses its own web index. As of May 2022, it covered over 10 billion pages and was used to serve 92% of search results without relying on any third-parties , with the remainder being retrieved server-side from the Bing API or (on an opt-in basis) client-side from Google.',
'Brave Search is a search engine developed by Brave Software, Inc., which is set as the default search engine for Brave web browser users in certain countries. Brave Search is a search engine developed by Brave Software, Inc. and released in Beta in March 2021, following the acquisition of Tailcat, ...'
]

Some of Brave Search's paid API plans allow snippet information to be used for AI learning, and some can store snippet information locally. Based on these facts, Mr. Ivanovs points out that ``Brave Search collects copyrighted information and sells it to others for a fee.''

Website administrators can refuse information collection by crawlers by editing the contents of 'robots.txt' and specifying the crawler's user agent. However, Brave Search does not publish user agents. For this reason, Mr. Ivanovs also sees the problem that information collection by Brave Search cannot be selectively eliminated.

After Mr. Ivanovs published his remarks on Brave Search, Mr. Josep M. Pujol, chief of Brave Search, said, 'Brave has the right to monetize the output of Brave Search.' Sites that refuse to be indexed by Google or sites that block Googlebot are designed not to be crawled.' However, Ivanovs said, ``Apparently Brave Search has the right to use copyrighted content for monetization just because it is a search engine,'' criticizing Brave Search's response. I'm here.

In addition, the chat AI ``ChatGPT'' also has a ``web browsing'' plug-in that searches the Internet and displays information, but in July 2023, it was reported that ``you can read paid articles for free.'' function has been suspended.

From the report that ChatGPT's web browsing function is suspended and you can read paid articles for free - GIGAZINE

However, ChatGPT publishes the crawler's user agent on the following page, so you can explicitly block information collection by editing 'robots.txt'.

Bots - Open AI APIs
https://platform.openai.com/docs/plugins/bot