The India AI Impact Summit 2026, convened by the Ministry of Electronics and Information Technology under the IndiaAI Mission within the Digital India programme, will be held at Bharat Mandapam in New Delhi starting Monday. This year’s edition is likely to be the biggest the country has seen so far in any city. With hundreds of foreign delegates, including top tech CEOs, heads of states and business leaders, already in Delhi for it, the city’s hold over artificial intelligence is as much in focus as are the topics of urban governance which will be discussed in the next five days.
When Sundar Pichai, Sam Altman, Demis Hassabis and Bill Gates discuss the theme of “Democratising AI” under the summit’s three pillars of People, Planet and Progress, they would be signalling both ambition and geopolitical intent.
The spectacle is impressive. The articulation is confident. Yet summits, however meticulously designed, cannot by themselves recalibrate a city’s lived reality. They may designate direction, but they cannot substitute for the execution. The question that must animate any serious assessment of the summit is not whether AI can transform the world in abstraction, but whether it can be domesticated into Delhi’s administrative fabrics in ways that are precise, accountable and socially equitable.
Public health
Public health is perhaps the most immediate domain in which the dividends of anticipatory intelligence could be realised. Delhi’s public hospitals, dispensaries and mohalla clinics carry a copious and fluctuating patient load. The city’s population now exceeds 1.64 crore, yet the hospital bed-to-population ratio remains around 2.7 per 1,000, well below the World Health Organisation’s recommended 5 per 1,000. Government facilities include about 38 multispecialty hospitals, 546 mohalla clinics, 427 dispensaries and other units. Mohalla clinics alone saw 1.94 crore patient visits in 2023, but visits dropped to 1.39 crore in 2024 due to drug shortages and supply issues.
“There are significant regional variations in genetic profiles, disease risks, diet and environment-led health conditions, infectious disease burden and lack of availability of electronic health records over time that make it difficult to build reliable models with western datasets alone; otherwise, it may lead to challenges like misdiagnosis, overdiagnosis, systemic bias, poor resource allocation and miscalibrated actions on the public health front,” said Nidhi Mathur, venture partner at Axilor.
Existing datasets are important starting points that can help us understand a lot about building the model to suit our needs. They have to be systematically supplemented with relevant data that will make them reliable for use in a regional context, but there is a lot we can learn, especially when biological data is involved, about the nature of data and the challenges we are likely to face in model building while collecting data in parallel.
In this era of rapidly advancing technologies, compressing technology development timelines will be key to staying competitive, she added. AI is already being operationalised globally to alleviate systemic stress in public health systems. During the Covid-19 crisis, hospitals in China deployed machine-learning tools to analyse CT scans at scale, accelerating diagnosis, standardising screening protocols and relieving clinician workload during peak surges.
“While AI-based solutions usually get a lot of attention, we have seen the role AI played during the pandemic that made a very clear case for the need for disease surveillance, building early warning systems and resource planning. Rising public healthcare costs, coupled with the need to improve healthcare access, make it imperative to think of AI as an integral part of the decision support system for policymaking,” she said.
Beyond emergency imaging, research platforms such as Health Sentinel have demonstrated how machine learning can mine news streams, epidemiological bulletins and surveillance data to detect outbreak signals in near real time. In India, the Translational AI for Networked Universal Healthcare (TANUH) Centre at IISc-Bengaluru represents a more structural intervention, seeking to equip frontline health workers with AI-driven decision-support systems for early detection, triage and risk stratification of major diseases. Together, these initiatives illustrate that AI’s utility in health is not confined to automation.
In Delhi, preliminary steps towards digital consolidation are visible, though still embryonic in scope. The Delhi government rolled out the Health Information Management System (HIMS) on July 24 last year, which enables online OPD bookings in 35 government hospitals, reducing queue congestion, paperwork redundancies and transaction friction for both patients and administrators.
“Start-ups are often better positioned than public institutions to experiment, build, iterate and deploy quickly. But in healthcare, both of them have to work together to build verifiable, privacy-sensitive and regulatory-compliant solutions. They attract entrepreneurs who are motivated to build systems that not only deliver affordable and accessible healthcare systems but also create viable commercial enterprises in the process,” she said.
Yet, while infrastructural expansion is underway, the systematic integration of AI—whether in predictive demand forecasting, automated triage support or supply-chain optimisation—remains to be institutionalised. The opportunity before Delhi is, therefore, not merely digital augmentation, but the construction of an integrated health intelligence architecture capable of moving from reactive burden management to calibrated, data-informed foresight. Two of the capital’s premier institutions—the Indian Institute of Technology and the All India Institute of Medical Sciences—took the first step in this direction when they decided to sign a memorandum of understanding in June last year to set up a Center of Excellence for AI in Healthcare.
Education
Education constitutes a second domain in which AI’s allure must be balanced by ethical sobriety. Internationally, AI tutors personalise content, automated grading systems accelerate evaluation, and learning analytics flag at-risk students. These tools promise efficiency and tailored support. Yet uncritical adoption might be dangerous. Excessive surveillance may engender anxiety rather than curiosity. Algorithmic labelling risks calcifying inequality. Data misuse can corrode trust in public institutions.
Delhi’s higher education institutions—IIT, Indian Institutes of Information Technology, University of Delhi and the Delhi Skill and Entrepreneurship University—have invested substantially in AI research and data science curricula. Technical supply is keeping pace with developments in AI. Rajeev Kumar, Dean of international programmes in the department of Computer Science and Engineering at Shiv Nadar University, opines that to make students AI-ready for the future, universities need to provide skills to students in all fields of study, whether it’s literature or computer engineering. He said the students should be conversant with the terminology and the system as a whole. In today’s time a framework needs to be created under which a student understands what the dos and don’ts of technology, particularly AI, he said.
The speed at which AI has developed in the last 5 years, particularly after ChatGPT has outdone everybody, including the AI experts. We have to educate ourselves with the technology, and a summit like this is happening at the correct time that will enable students and faculty members to be aware of AI. This is because AI now has come to the point where it can’t be ignored, and one needs to be updated with every latest development. “The government should help formulate a training programme for the teachers of this country to make them fluent in AI. Not all universities are equipped with the technology and funds needed to integrate AI into the curriculum, and the gap needs to be bridged,” he said.
The governance framework, however, remains embryonic. A coherent public policy must precede large-scale deployment in classrooms. Teachers must not be relegated to interpreters of opaque dashboards. “The universities abroad are bringing AI policies into the classrooms where the allowed usage of AI in every course is designed. The policies are evolving, and as AI progresses in the next few years, the policies need to develop further,” he said.
A piece of information generated by AI needs to be looked at very critically, and one must check all possible sources, he said, adding that the job of the educators is to train the minds of students to question the answers and not blindly trust the answers generated by the AI system.
Traffic and pollution
Transport and mobility present a more immediately tangible arena for AI’s pragmatic utility. Delhi’s congestion is chronic and measurable. Signal desynchronisation, erratic bus intervals, illegal parking and corridor bottlenecks constitute gridlock. Other cities offer instructive precedents. Delhi has already adopted AI cameras for challan generation. Yet isolated instruments, however sophisticated, cannot transform systemic flow unless they dovetail.
Nakul Anand, a veteran in AI Strategy & Innovation and Platform Architect at Decimal Technologies Pvt Ltd said, “In urban planning, the sectors most impacted by AI include traffic management, pollution monitoring, disaster preparedness, and utilities like water and energy distribution. These areas generate continuous real-time data, making them suitable for AI-driven analysis”. Predictive models allow cities to move from reacting to problems to anticipating them.
This helps governments design policies, allocate resources, and monitor outcomes more effectively. From an architecture perspective, the impact is strongest where data systems across departments can be integrated securely and reliably, he said. Examples showed us that predictive modelling of bus demand can smooth intervals and reduce overcrowding.
The objective is not velocity for its own sake but reliability—the conversion of chaotic motion into well managed circulation. Anand highlighted that adaptive traffic control replaces fixed-timer traffic lights with systems that adjust signal timing based on live camera and sensor inputs. He said, “In a city like Delhi, with very high vehicle density, this can reduce bottlenecks”.
However, the system requires reliable sensor networks, enforcement, and a centralised command platform, he added. Emphasing that governance, maintenance, and data quality are as important as the algorithms themselves, he added, AI can improve traffic flow within one to two years if implemented properly. Pollution control usually takes longer because air quality depends on multiple factors beyond traffic alone.
“AI contributes mainly through prediction, scenario modelling, and targeted interventions. The timeline depends on how quickly cities can integrate data systems and institutional workflows.”
Air quality remains Delhi’s most pestering affliction. Current responses under the Graded Response Action Plan often rely on citywide averages and blanket prohibitions. While occasionally necessary, such measures are blunt and imprecise. Artificial intelligence offers precision.
AI Producer and AI Consumer
At the national level, technological sovereignty looms large. Delhi, by virtue of academic density and policy proximity, could anchor a national AI grid — a compute commons under public aegis accessible to researchers and start-ups. Stating that we do not need to start from scratch. India already has large volumes of health data from hospitals, labs, insurance claims, and telemedicine platforms, Anand said,
“The challenge is that this data is fragmented and not always structured for AI use.” He said, “The focus should be on standardising, integrating, and validating existing data, while collecting better-quality data in areas such as lifestyle diseases and regional language records. This approach is consistent with how AI platforms are built in regulated sectors, where interoperability and auditability are critical.”
Transparent procurement, open datasets and interoperable standards would fortify sovereign capacity and prevent overreliance on external platforms. AI is not destiny; it is a policy. It can entrench asymmetry or foster inclusion. The India AI Impact Summit 2026 will doubtless generate effusive speeches, convivial exchanges and strategic communiques. Yet discourse, however eloquent, is fleeting without execution.
Delhi stands at an inflection point. It may allow AI to remain a flourish of aspiration, a symbol of technological modernity invoked with gusto or it may embed AI within the machinery of governance, labour policy and civic systems.