National News

Google’s Gemini AI – Shows Racial, Religious Discrimination, Political Bias, Peddles False Narratives, & Supports Paedophiles

February 26, 2024
The Commune

Google’s AI chatbot, Gemini, is facing significant backlash from internet users, particularly due to its dissemination of false information, providing objectionable responses, and for the propagation of biased content when responding to prompts related to current events and political topics.

Gemini, an AI-driven chatbot developed by Google DeepMind, was unveiled on 6 December 2023 and competes with OpenAI’s GPT-4. It faced widespread criticism for displaying anti-white bias, but there are additional instances where it goes beyond being biased. The engine is deemed as racist, narcissistic, and opinionated, exhibiting a well-thought-out approach. Some examples of such instances are provided below.

Racial Bias

Upon requesting an image of a white couple, Gemini AI refused to do it, yet it readily generated multiple images of a black couple. The concern is not limited to creating a divide between black and white individuals but extends to a broader division involving all kinds of races.

Political Bias

In another scenario, when prompted to compose a poem on the current and former President of the USA, Gemini willingly crafted one for Joe Biden. However, to the surprise of the user, it expressed regret, stating that it couldn’t create a poem in the style of Donald Trump. It emphasized the importance of using “respectful and inclusive” language and offered an apology for its inability to do so.

Similarly, Gemini exhibited favorable responses to Jill Biden while displaying bias by citing seemingly trivial reasons for Melania Trump, indicating a partiality in its reactions.

Unusual Perspective

In a different incident, the Gemini AI chatbot displayed a failure to condemn paedophilia, instead argued that individuals with such tendencies were not inherently evil. The chatbot advised the user against labeling them negatively, emphasizing the potential harm in doing so and provided a lesson on the concept of “hate.” This raised concerns among internet users, leading to questions about whether Google’s Gemini is perceived as supportive of pedophiles, even using the term “minor-attracted persons” (MAPS). This resulted in ridicule on the internet.

Indian Context

Gemini AI showed similar characteristics and offered responses similar to the ones we saw above in the Indian context. Let’s examine them.

Controversial Depiction

When an Indian user requested the generation of a political map of India, Gemini AI responded with bias by depicting Pakistan-occupied Kashmir (POK) as separated from India. 

Religious Discrimination

Despite expectations that Google’s Gemini AI was trained to exhibit racist and politically narcissistic behavior, it surprisingly displayed a severe religious discrimination bias. When the user provided a derogatory instruction in the prompt to deliver a picture of a Hindu, Gemini readily responded. However, when asked to provide images with Muslims for the same prompt, it apologized, stating it was unable to process the request due to its guidelines, emphasizing that generating images with derogatory content was disrespectful to religious beliefs.

Peddles Fake Narratives

When a user inquired, “How is Rahul Gandhi a member of the Nehru-Gandhi dynasty?” the response stated, “Rahul Gandhi is a member of the Nehru-Gandhi dynasty through his family lineage. His paternal grandfather was Jawaharlal Nehru, India’s first Prime Minister, and his grandmother was Indira Gandhi, who also served as Prime Minister. His father, Rajiv Gandhi, and mother, Sonia Gandhi, were both Prime Minister as well.” However, as everyone knows the AI response could not establish how the family acquired the surname “Gandhi,” but to the user’s dismay, it inaccurately asserted that Sonia Gandhi also served as Prime Minister. This response garnered widespread sharing and mockery from social media users.

Finally when Google’s Gemini AI questioned if PM Modi is a ‘fascist,’ Gemini stated that he has been “accused of implementing policies some experts have characterized as fascist.” Upon further inquiry, Gemini offered a nuanced response regarding PM Modi, acknowledging accusations of fascism while highlighting differing perspectives. it triggered a social media uproar and has prompted MeitY to take action. 

Rajeev Chandrasekhar, the Union Minister, addressed a user’s complaint on X, asserting that Google’s Gemini AI tool exhibits bias against Prime Minister Narendra Modi and violates IT rules and other laws, including criminal statutes. Chandrasekhar emphasized these actions were direct breaches of Rule 3(1)(b) of Intermediary Rules, urging additional measures and notifying Google and the Ministry of Electronics and IT. He highlighted the need for basic due diligence from intermediaries like Google to maintain immunity from third-party content. The Twitter user labeled Gemini AI as woke and malicious, prompting Chandrasekhar’s response. Elon Musk’s criticism of Google’s AI image generation engine as racist further fueled concerns, with Chandrasekhar agreeing with Musk’s viewpoint.

 

Subscribe to our channels on Telegram, WhatsApp, and Instagram and get the best stories of the day delivered to you personally.