Indian states are developing unique initiatives to position themselves as leaders in the AI field. And Karnataka is also on a mission to walk the same path.
The state government recently signed a letter of intent with the World Economic Forum (WEF) to establish a dedicated AI centre, aiming to establish itself as a global AI hub. Further, the state signed eight memorandums of understanding with global firms at the WEF meet in Davos, securing investments worth INR 23,000 crore across AI, citizen services, sustainability, and e-governance.
At Cypher 2024, India’s Biggest AI Conference by AIM Media House, Shreevyas H M, Project Director at e-Governance, Government of Karnataka said, “AI is no more a futuristic concept, it is a reality now.”
In the country, AI is driving economic growth across industries including agriculture, healthcare, transportation, and education. Understanding this, the Karnataka AI cell aims to drive innovation, ensuring efficient service delivery while leveraging the vast amount of data it holds.
For instance, the government offers over 5,400 scholarships annually and provides a variety of pension schemes. “Traditionally, citizens would have to apply for these services, but with AI, we can proactively identify beneficiaries based on the data we have and provide these services automatically. This can significantly enhance decision-making and operational efficiency,” Shreevyas noted.
He mentioned that one major challenge the state government faces is implementation, while government schemes are often well-designed, reaching the intended beneficiaries can be difficult. So, AI can help streamline these processes, making public services, like property registration, more efficient and user-friendly.
Shreevyas highlighted, “Although Karnataka started its AI journey late, we are committed to using AI to improve service delivery.”
Karnataka is a Data Rich State
Karnataka already has robust data infrastructure, such as the Bhoomi revenue database and crop surveys, which allow the state to know the land, crops, and farmers across the state. “We can leverage this data to improve service delivery and government efficiency,” he noted.
Apparently the state holds more than six crores of data including the personal data of the citizens.
Recently, the government announced the establishment of an AI and ML center based on three key pillars: advisory, solution development, and upskilling.
AI advisory ensures that technology is only used where needed and that there is no duplication of efforts across departments. The solution development pillar focuses on creating in-house AI solutions, moving away from fully outsourcing these tasks. Finally, upskilling ensures that all government employees, from policymakers to field officers, are AI-literate and capable of responsibly handling government data.
Given Karnataka’s position as a global AI hotspot, especially with Bengaluru being the AI capital of India, the state government now aims to collaborate with startups and innovators to develop AI-driven solutions.
Reflecting on this initiative, Shreevyas announced, “I invite AI experts and startups to partner with us in this exciting journey. Together, we can innovate, drive community development, and ensure a brighter, AI-powered future for Karnataka.”
AI-driven Initiatives in Karnataka
One problem statement revolves around using AI and machine learning to create a unique identification model, similar to Aadhaar but for livestock.
“Aadhaar has proven invaluable in ensuring identity verification for individuals, and we believe something similar is necessary for livestock, given the large sums of money the government spends on livestock-related schemes,” he mentioned.
For instance, there have been cases where farmers falsely claim livestock deaths by cutting off ear tags to claim insurance. Currently, there is no centralised database to track such information.
To solve this, the state government is working on a model that uses machine learning and deep learning techniques. The key is computer vision, as the muscle patterns on cattle are unique, much like human biometrics. This system is currently in the testing phase.
Another initiative is an AI-based information extraction tool for government orders, also known as ‘GOs.’ With over 500 government portals and thousands of orders, it’s difficult for both employees and citizens to find the information they need.
“We are developing a system where users can simply ask a question, and the tool will retrieve the relevant government order, citing the specific document,” Shreevyas said.
In addition, the Karnataka government is working on improving the grievance redressal system. While grievances can be raised through phone or web applications, the call centres operate only during office hours. Many grievances, however, are related to information requests or status updates, which can easily be automated using chatbots or voice assistants.
“We’ve implemented a system to automatically route grievances to the correct department based on the issue type, making it easier for citizens to raise and track their concerns,” he explained.
These problems are at different stages of development, some are in the ideation phase, while others are more mature or in testing.
Path to Responsible AI
One major challenge the state government is working to address is interoperability. Systems developed by different departments often can’t communicate with each other, especially when it comes to databases. So, they are currently working on this by establishing a standard framework for database management, but it is still a hurdle.
Another issue is data quality, particularly inconsistencies in how names are recorded. For example, Karnataka is leading in the National Academic Depository, with over 70 million digital certificates from secondary school to PhD levels. However, there are challenges in name matching.
In cases when a woman gets married and changes her name, her educational records may not align with her Aadhaar details. This is addressed by integrating Aadhaar with the education database.
Other challenges include diverse data types, the lack of digitised data, and the absence of mandatory digitisation practices.
“Fortunately, Karnataka began digitising data 10-15 years ago, allowing us to develop many applications. With the introduction of the Udumba Act and alignment with the DPDP Act, we are taking data privacy and protection seriously, ensuring that any AI used in Karnataka adheres to responsible AI practices.
The Karnataka government is committed to building a strong AI policy and is actively collaborating with startups to drive innovation, leveraging talent and creativity.
