Baidu denies Japanese government’s accusations against its Japanese input method

Don’t be evil. Baidu claims its Japanese input method does not upload any personal information onto its servers, after a Japanese government agency suggested to governmental institutions and universities that they should avoid using the input methods of Baidu, Google as well as Microsoft when dealing with classified documents. The Japanese input method of China’s search giant started off at the end of 2009.

Baidu’s press release to Huxiu in response to Japan’s negative reviews:

Baidu’s Japanese language input method employs cloud technology that is widely used by similar products worldwide. This technology, together with natural language processing technologies, provides our users with a constantly updated, cloud-based dictionary. Encrypted information is uploaded to our servers and processed as aggregated data, and as such cannot be linked back to any individual users (for instance, through account names). Therefore, this technology does not pose any risk of information leaking to our users. The servers for Baidu’s Japanese language input method and the data consequently generated are all located locally in Japan, and adhere to the country’s laws and regulations.

Baidu’s Japanese language input system for mobile devices was created by a software development team that Baidu acquired in December 2011. This input system was among the first in Japan to use this type of cloud technology, and it has been well received by our users in Japan for providing a great user experience.

We are currently communicating with relevant Japanese government agencies to reach a better, mutual understanding on the matter at hand. Baidu Japan has always respected and operated under Japanese law. As such, our Japanese language input methods have always been in compliance with Japanese law, as well as with relevant international practices.

Baidu attaches great importance to transparency and the protection of our users’ information. We are actively working to strengthen the information disclosure mechanisms in future versions of our products to provide our users with a clearer understanding of how our products work, so that they can enjoy using them worry-free.