"AI bias can manifest in ways not obvious in Western contexts": PM Modi calls for global cooperation on AI
New Delhi, February 17
Prime Minister Narendra Modi highlighted concerns about bias and limitations in Artificial Intelligence, noting that AI systems can reflect gender, language, and socio-economic biases and may impact India differently due to its vast linguistic and cultural diversity.
The Prime Minister said that as AI adoption increases, the risks also grow, stressing the need for global cooperation to address bias.
In an interview with ANI, PM Modi said, "The concerns regarding bias and limitations in AI remain deeply relevant. As AI adoption accelerates, the risks also scale. AI systems can inadvertently perpetuate biases related to gender, language and socio-economic background. The AI Impact Summit 2026 is bringing together various stakeholders and creating global awareness on matters such as biases and limitations of AI. This is an issue that needs global cooperation."
He noted that AI models trained mainly on Western or English-language data may not work effectively for rural users or speakers of regional languages in India, and emphasised the importance of developing India-specific, inclusive AI systems.
"For India specifically, we face unique challenges and opportunities. Our diversity, linguistic, cultural, and regional, means that AI bias can manifest in ways that might not be obvious in Western contexts. An AI system trained primarily on English data or urban contexts may perform poorly for rural users or speakers of regional languages," the Prime Minister said.
The Prime Minister said that India is beginning to address concerns around Artificial Intelligence bias more systematically, highlighting efforts to create diverse datasets, promote AI development in regional languages, and expand research on fairness and bias in academic institutions and technology companies.
"The positive development is that India is beginning to address this more systematically. We're seeing increased focus on creating diverse datasets that represent India's plurality, greater emphasis on AI development in regional languages, and growing research on fairness and bias in Indian academic institutions and tech companies," he said.
A day earlier, Prime Minister Modi inaugurated the India AI Impact Summit 2026 at Bharat Mandapam in the national capital, underscoring India's commitment to responsible and inclusive Artificial Intelligence.
The India AI Impact Summit 2026 strengthens India's role as a key platform for shaping the global AI agenda. Anchored in the Seven Chakras and the Three Sutras of People, Planet, and Progress, the Summit advances a development-oriented framework for artificial intelligence.
— ANI
Reader Comments
Finally, someone said it. Most AI tools assume a nuclear family, urban lifestyle, and English fluency. They fail in our joint family setups, local market contexts, and regional festivals. We need our own datasets, not just translated Western ones.
Global cooperation is key, but action is needed here first. I hope this summit leads to real funding for Indian language NLP projects and mandates for bias testing in government-tendered AI systems. Talk is good, implementation is better.
As someone working in tech, this resonates. The bias in image generation AIs is stark—they often can't generate accurate traditional Indian attire or settings. PM Modi is right; diversity in training data isn't a luxury, it's a necessity for fairness.
Good initiative. But I respectfully hope the focus on 'India-specific' AI doesn't turn into a protectionist wall. The goal should be to feed our diversity *into* the global AI pool, making it richer for everyone, not just creating a separate silo.
Absolutely! An AI trained abroad might not understand the socio-economic nuance of a 'kirana' shop vs. a supermarket, or the importance of regional crop patterns. Our context is everything. Jai Hind! 🙏
We welcome thoughtful discussions from our readers. Please keep comments respectful and on-topic.