Google Translate was found to have a gender bias issue, where it often defaults to masculine pronouns and stereotypes when translating text, prompting accusations of "sexism".

Historically, Google Translate often defaulted to masculine translations for pronouns and certain professions and adjectives. For example, it would translate "doctor" as male and "nurse" as female, reflecting societal biases present in its training data, which consists of millions of previously translated texts.

As users and researchers observed, this can lead to inaccurate and sexist translations, perpetuating existing biases and stereotypes. The issue is not limited to pronouns, as the algorithm also tends to associate certain professions and roles with specific genders. 

In response, Google updated its system to provide masculine and feminine translations for gender-neutral words in several languages, including French, Italian, Portuguese, and Spanish. The feature allows users to see both forms for terms like "surgeon" when translating from English. 

Google said it would also plan to expand these gender-specific translations to more languages and address non-binary gender representations in future updates.

Despite these improvements, some questioned the effectiveness of Google's approach, arguing that that while dual translations are a step forward, they still do not fully address the underlying biases in the data or the algorithms used.

 The challenge lies in the fact that machine learning models can perpetuate biases present in their training datasets, and merely providing two options does not eliminate the inherent biases in how gender is represented in language.