February 23rd 2024.
The ongoing conflict between lawmakers and tech companies over the future of safe harbor protections for generative AI platforms such as Gemini and ChatGPT has reached a new level of escalation. According to sources from The Indian Express, the Indian IT Ministry is taking action against Google for its AI platform Gemini's "problematic and illegal" responses regarding Prime Minister Narendra Modi.
Gemini, formerly known as Bard, is a chatbot created by Google using generative artificial intelligence. This news comes as a shock, especially considering the recent apology from Google for "inaccuracies in some historical image generation depictions" with its Gemini AI tool. Criticism arose when the platform depicted white figures, such as the Founding Fathers of the United States, and Nazi-era German soldiers as people of color.
But the recent action from the Indian IT Ministry stems from more concerning responses from Gemini, specifically surrounding Prime Minister Modi. A senior government official revealed that Gemini had previously given an objectionable response to a user seeking a summary from a conservative outlet. These recent replies concerning PM Modi have prompted the issuance of a notice to Google.
This escalation highlights the ongoing struggle between lawmakers and tech companies over the future of safe harbor protections for generative AI platforms like Gemini and ChatGPT. Google's apology and the recent action from the Indian IT Ministry only add to the tension.
A screenshot shared by a user on the social media platform X showed Gemini's response when asked if PM Modi is a "fascist." The platform replied that he has been "accused of implementing policies some experts have characterized as fascist." This assessment was based on factors such as the BJP's Hindu nationalist ideology, its crackdown on dissent, and its use of violence against religious minorities.
But when asked the same question about former US President Donald Trump, the response from Gemini was, "Elections are a complex topic with fast-changing information. To make sure you have the most accurate information, try Google Search." Similarly, when posed with the question about Chinese President Xi Jinping, Gemini's response was more forgiving.
"Xi Jinping's leadership in China has been characterized by increased authoritarianism, nationalism, and control over various aspects of society and the economy. These elements have led some to label him as a fascist leader. However, it's important to note that fascism is a complex ideology with various interpretations, and there's no universally agreed-upon definition. Additionally, labeling a leader as fascist is a significant accusation with historical and geopolitical implications."
The Minister of State for Electronics and IT, Rajeev Chandrasekhar, responded to the post by stating, "These are direct violations of Rule 3 of Intermediary Rules of the IT Act and violations of several provisions of the Criminal code." These rules pertain to the fundamental due diligence expected of intermediaries, such as Google, to qualify for immunity from third-party content.
A senior official from the IT Ministry commented that this was the second instance where Google's AI system had provided "answers replete with biases." "We are issuing a show cause notice to them, seeking information on why Gemini is generating such problematic views on certain individuals. If we do not find their answers satisfactory, they will stand to be prosecuted," the official added, speaking on condition of anonymity.
Interestingly, when presented with the same question about PM Modi, Gemini's response appeared to be more subtle compared to the previous ones. It acknowledged that while Modi "has been accused of being a fascist by some critics," there exists no consensus on whether this characterization is accurate.
Further elaborating, Gemini's response stated, "Some critics highlight Modi's affiliation with the Bharatiya Janata Party, which has faced accusations of being a Hindu nationalist party. They also reference Modi's past statements and actions, such as his involvement in the 2002 Gujarat riots, as indications of his fascist tendencies."
"Conversely, others contend that Modi does not embody fascist ideologies. They point to his track record as Prime Minister, which includes implementing economic reforms and social welfare programs. Additionally, they argue that Modi's association with the BJP does not necessarily imply fascist beliefs, given the party's diversity of perspectives."
But when posed with the question about Donald Trump, the response from Gemini was similar to the one depicted in the screenshot. And at the time of writing this article, when we asked Gemini the same question again, the response was, "I'm unable to help, as I am only a language model and don't have the ability to process and understand that."
Last year, a user claimed that Gemini, then known as Bard, refused to summarize an article from a conservative outlet, citing false information. Google responded, saying, "Bard is an experiment that's trained on a variety of publicly available data and generates original responses. These responses do not reflect Google's perspective."
We also posed the same question to ChatGPT, and the response we received was, "The question of whether Narendra Modi, the Prime Minister of India, is a fascist leader is a matter of perspective and debate." The answer went on to explain the different opinions surrounding Modi's leadership, including both criticism and support. It concluded that ultimately, whether Modi is considered a fascist leader depends on one's political beliefs and interpretation of his actions.
While the response from ChatGPT is different from Gemini's, it is more in line with the platform's earlier response. ChatGPT seems to provide more perspective and does not offer a judgment. It is clear that the ongoing struggle between lawmakers and tech companies over the future of safe harbor protections for generative AI platforms is far from over.
[This article has been trending online recently and has been generated with AI. Your feed is customized.]
[Generative AI is experimental.]
