Google Autocomplete falsely associates Japanese man with crimes
Google Autocomplete falsely associates Japanese man with crimes
Occurred: March 2019
Report incident ๐ฅ | Improve page ๐ | Access database ๐ข
A Japanese man successfully sued Google for defamation after the search engine giant's Autocomplete function falsely associated him with criminal acts, resulting in his reputation being tarnished, the loss of his job and his inability to find further employment.
According to the man's lawyer, over 10,000 Google search results had wrongly associated his client's name with crimes committed by someone else with the same name, and that Autocomplete's algorithms had been directing users to potentially false or misleading information.ย
Google had reputedly refused to remove the offending words from its system when the man had first complained about them to the technology company. The judge ordered Google to delete the terms and to modify part of Autocomplete in Japan.ย
Google had already lost a lawsuit against the same man and refused to delete the relevant information, having claimed it was not subject to Japanese law and that its Autocomplete system could not violate privacy as it is automatically generated and depended on what was already available online.
Autocomplete
Autocomplete, or word completion, is a feature in which an application predicts the rest of a word a user is typing. In Android and iOS smartphones, this is called predictive text.ย
Source: Wikipedia ๐
Operator: Alphabet/Google
Developer: Alphabet/Google
Country: Japan
Sector: Business/professional services
Purpose: Predict search results
Technology: NLP/text analysis; Deep learning; Machine learning
Issue: Accountability; Accuracy/reliability; Mis/disinformation
https://abcnews.go.com/blogs/technology/2012/03/court-tells-google-to-suspend-autocomplete/
https://www.bbc.co.uk/news/technology-17510651
https://www.theregister.com/2012/03/26/google_autocomplete_japan/
https://www.theregister.com/2012/06/19/google_japan_defamation_autocomplete/
Page info
Type: Incident
Published: March 2019