.avif){kind=small}
Google's autocomplete feature faced legal challenges in France and Japan for associating public figures and individuals with sensitive or defamatory terms. In France, organizations sued over the systematic association of prominent names with the term 'Jewish', alleging the creation of an illegal ethnic database. In Japan, an individual successfully sought an injunction after the autocomplete feature linked their name to crimes they did not commit, causing significant professional harm.
The organizations SOS Racisme, Union of Jewish Students of France, Movement Against Racism and for Friendship Among Peoples are suing Google due to its autocomplete software suggesting "jewish" when the names of certain public figures were searched on the platform.
The autocomplete algorithm systematically associated the term 'Jewish' with prominent public figures, which French plaintiffs argued created an illegal ethnic file and represented an anti-Semitism problem.
The autocomplete algorithm generated unexpected and harmful associations post-deployment, which was an unintentional outcome of its search prediction goal.
Minimal or No Risk: The autocomplete feature is a general search utility that poses low inherent risk to society, similar to spam filters or simple applications, and does not fall under high-risk categories like critical infrastructure or law enforcement.
Severity ratings are based on this scale.
Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.
Reported: Yes, the report describes financial losses resulting from the autocomplete associations.
Indirectly caused: The unnamed man in Japan suffered a sudden loss of his job and was subsequently rejected by several companies when applying for new jobs due to the autocomplete feature linking his name to crimes.
Inferred additional harm: The Japanese man likely suffered substantial lost wages and career damage, and the French companies in the 2009 lawsuit likely suffered business and revenue losses due to being associated with the word 'scam'.
Reported: Yes, the report describes the spread of objectionable or sensitive content through autocomplete suggestions.
Directly caused: Google's autocomplete feature directly generated and displayed sensitive and potentially defamatory terms like 'Jewish' and 'scam' alongside names of individuals and companies to search users.
Inferred additional harm: Many other users globally were likely exposed to biased, toxic, or defamatory autocomplete suggestions generated by the algorithm based on search trends.
Reported: Yes, the report describes differential treatment based on sensitive characteristics.
Directly caused: The autocomplete feature systematically associated prominent public figures with the term 'Jewish', which French plaintiffs argued created an illegal ethnic file, treating individuals differently based on perceived religious or ethnic background.
Inferred additional harm: The systematic categorization of individuals by ethnicity or religion in search results could lead to broader societal discrimination and bias against Jewish public figures.
Reported: Yes, the report describes violations of civil rights and legal protections.
Directly caused: The creation of an 'ethnic file' through autocomplete is outlawed in France as a violation of civil rights and anti-discrimination laws.
Inferred additional harm: The automated generation of ethnic databases compromises the fundamental right to non-discrimination and equal treatment under French law.
Reported: Yes, the report describes privacy-related concerns.
Directly caused: The autocomplete feature compiled and publicly associated sensitive personal data (ethnicity/religion) with individuals' names, which plaintiffs characterized as an illegal ethnic file.
Inferred additional harm: The algorithm potentially infers and exposes private, sensitive, or non-public information about individuals' personal lives and backgrounds without their consent.
Reported: Yes, the report describes the generation of false and misleading information.
Directly caused: Google's autocomplete algorithm falsely connected a Japanese man's name with crimes he had not committed.
Inferred additional harm: Users searching for these individuals or companies were misled into believing false claims (e.g., that a person committed a crime or a company was a scam), eroding trust in search results and factual reality.
Google autocomplete algorithm generated defamatory and sensitive search suggestions in France and Japan, leading to legal challenges over illegal ethnic categorization and personal reputational damage. The incident represents a minor societal stability and human rights concern managed through standard legal and regulatory frameworks.