India AI Mission: Parliament Data MoU for Indigenous LLMs
Why in the news
The India AI Mission tied up with Parliament for data to train home-grown LLMs, while the government spelled out its plans for compute and self-reliance.
Key facts
- Data sources: Parliament, Doordarshan and All India Radio.
- Shared compute: 14,000 GPUs.
- Target: domestic GPU manufacturing within 3-5 years.
- Two competing instruction set architectures are being assessed for indigenous GPUs; experts will decide.
Policy stance
- The Minister warned that open source can stop being open later, so proprietary Indian models get priority.
- The government will partner industry, professors and startups.
- After 100 university labs for 5G training, AI skilling will follow a similar model.
- On the India-US partnership, Mr. Vaishnaw stressed trust, respect for IP rights and decade-long co-development.
Significance
- India should be a producer of cutting-edge AI, not only a consumer.
- Indigenous models plus compute build strategic autonomy and competitiveness.
Exam angle
- Compute: 14,000 GPUs.
- LLM = Large Language Model.
- MoU partner: Parliament.