1 day ago
1
A radical of researchers have travel up with the ‘Next-Billion AI Index: The compass for AI inferior and adoption successful the planetary majority’ .
This caller survey challenges the planetary AI industry’s obsession with benchmark rankings with its caller framwork Next Billion AI Index Framework (Nexbax AI Index). It argues that the metrics dominating AI improvement successful the satellite are fundamentally useless for the billion-plus users crossed India and the Global South.
The Nexbax AI scale is developed by Ambrish Rawat, Kush R. Varshney and Jessica He from IBM research, Subhabrata Majumdar from Indian Institute of Management Bangalore (IIMB), Claudio Pinhanez from University of Sao Paulo, Yann Le Beux from YUX Design, Satyapriya Krishna from Harvard University, Rahul Gupta from Amazon and Rumman Chowdhury from Humane Intelligence.
Even successful India’s AI argumentation discourse, NITI Aayog progressively recognises that readiness depends connected affordability, multilingual access, and compliance-by-design. According to the researchers, the Nexbax AI Index gives that intuition a rigorous and a measurable form.
Subhabrata Majumdar, adjunct prof from IIMB, said, “This model is developed with coauthors not lone from India but radical spanning from crossed Africa, Europe, and North America. We are trying to measure if the existing AI systems tin beryllium deployed, trusted, and sustained nether the constraints that specify adoption successful markets similar India, due to the fact that Indian marketplace has divers users utilizing devices with intermittent connectivity, low-cost devices, multilingual and low-literacy users, and informal small-business workflows.”
He said, “The task was initiated astir six months agone during the India AI Summit successful February 2026. The thought down processing the model was to usher the improvement of amended AI systems truthful that these systems tin lick the problems of radical moving astatine the grassroot level. For example, giving the astir important wellness updates to ASHA workers who enactment successful the distant villages of India wherever net connectivity hasn’t seeped successful yet. Such AI systems oregon developments indispensable beryllium brought successful India to assistance woody with the existent satellite challenges.”
The Nexbax AI Index is developed connected the ground of 10 measurable dimensions organised nether 3 themes: Effective Efficiency, Operational Practicality, and Societal Integrity. The 10 dimensions included outgo effectiveness and terms performance, assets efficiency, adaptability and customisability, interoperability, resilience, reliability and robustness, usability and automation, acquisition and empowerment, trustworthiness and ethics, multiculturalism, inclusivity and pluralism, openness and collaboration.
The squad conducted a formative adept valuation of the archetypal mentation of the scale done one-hour idiosyncratic interviews with eleven experts (participants) moving connected AI exertion and applications for lawsuit usage. These 11 participants included founders of assorted companies’ developers, and merchandise leaders.
The extremity of this valuation was to measure the clarity, actionability, and perceived accuracy of the dimensions for method and merchandise stakeholders, and to iterate connected them based connected adept feedback. Participants tested the scale by standing 3 illustrative exertion configurations crossed the 10 dimensions. They shared feedback connected their acquisition utilizing the dimensions, including wherever they saw value, points of confusion, and suggestions for improvement.
Prof. Majumdar explained, “Overall, the experts we interviewed felt that the scale was utile for reasoning astir AI inferior and adoption successful next-billion settings from a developer and merchandise perspective. One subordinate explained that the tools utilized to measure customers are beyond localisation, easiness of use, reliability, the cost, the acceptable up clip — each those things are hit. Those are the aforesaid factors that the founders see to prime their vendors. When asked to place the astir important decision-making considerations successful their work, 10 participants chose outgo effectiveness, 5 chose usability and automation, 4 chose trustworthiness and ethics. It is worthy noting that these dimensions span each 3 themes, providing aboriginal awesome that each taxable captures chiseled and meaningful aspects of real-world determination making.”
The squad recovered that from a usability standpoint, the experts mostly recovered the dimensions and rubrics casual to apply, though immoderate felt that it would person been hard without the guidance and clarifications they received from the researchers during the survey session.
In that sense, the worth of nexbax is not lone that it produces an index, but that it shifts the entity of valuation itself: from asking which AI systems are astir susceptible successful idealised settings to asking which systems make the conditions for useful, inclusive, and sustainable adoption wherever the adjacent cardinal users unrecorded and work.
The researchers person clarified that a aboriginal iteration of this enactment volition see aggregated, system-specific ratings from a larger and much divers stakeholder sample, including public-sector implementers, civilian nine organizations, section domain experts, and affected users.
