As artificial intelligence begins to shape how products are priced, how companies compete and how data is shared, economies everywhere are reaching a crossroads — and India is no exception. AI promises efficiency and innovation, but without the right guardrails it also threatens to tilt markets, favour the powerful and undermine trust.
At the heart of the challenge is a simple but unsettling question: what happens when algorithms, left to their own logic, start learning to collude? Competition lawyers and technologists say that preventing such “silent coordination” among AI systems is no longer theoretical.
Rahul Rai, partner, Axiom5 Law Chambers, said the risk lies in how algorithms are designed. The solution, he argued, is to build guardrails directly into their architecture.
Pricing systems, for example, should draw only on public information, not on the kind of private competitor data that can blur the line between insight and manipulation. Human oversight, he added, must remain part of every pricing decision, backed by regular audits.
Others in the technology community believe transparency is the missing link. Sagar Vishnoi, who runs the think tank Future Shift Labs as director and co-founder, said AI’s decision-making should be made visible to regulators and, where possible, to consumers.
“If an algorithm changes prices or recommends actions, someone needs to be able to explain why,” he said. Regular auditing, he added, should become as routine as financial reporting.
That combination of design discipline and openness is becoming central to how policymakers think about competition in the age of machine learning. Yet for India, where startups struggle for access to data and computing power, the challenge extends beyond policing the giants.
A question of access
For small companies trying to build their own AI models, the problem is not regulation but resources. The cost of training even modestly sized systems can be prohibitive, and valuable data remains concentrated in the hands of large firms or government silos.
There are efforts to change that. The government’s broader IndiaAI Mission, launched to build a national AI ecosystem, aims to make such resources more accessible. The initiative proposes shared computing infrastructure, data platforms and innovation centres across the country to support research and startups. If implemented effectively, experts say it could help translate policy intent into tangible capacity for smaller firms.
Alongside it, the National Data Governance Framework Policy seeks to create large, anonymised public datasets that developers can use to train models. Experts say that kind of initiative could give startups a fighting chance.
Both Rai and Vishnoi see potential in open data and open-source models to reduce dependence on costly private infrastructure.
Vishnoi believes public–private partnerships could help bridge the gap. Cloud computing subsidies, incubators and access to shared compute infrastructure, he said, could make innovation more democratic. “If India wants a vibrant AI ecosystem, small players need the same raw material as the big ones,” he said.
Rai adds that the government could go further by offering tax incentives or subsidies to reduce computing costs, ensuring access to AI does not become a privilege limited to a few cities or sectors.
AI is beginning to change how products are priced for individuals, a practice known as personalised pricing. While companies argue it allows them to tailor offers, critics worry it can lead to hidden discrimination, especially when customers have no way of knowing how or why prices change.
Regulating that kind of pricing, experts say, will test the limits of consumer protection and competition laws. Rai believes the focus should be on ensuring fairness, not micromanaging prices.
“Regulators must be careful not to replace the firm’s judgment with their own,” he said, warning that overreach could chill competition.
Vishnoi, on the other hand, sees transparency as the first line of defence. Regulators, he said, should require companies to explain how personalised prices are determined and allow consumers to flag unfairness through dedicated platforms.
Watching the giants
While regulators grapple with questions of fairness at the consumer level, an equally pressing challenge lies in how power is distributed among the firms building these systems.
As India’s AI economy matures, another risk looms: consolidation. Large technology companies have been buying up promising AI startups or securing exclusive deals for key technology. Regulators worry this could create data monopolies that suffocate smaller competitors before they can scale.
The concern is hardly unique to India. The 2025 Nobel Prize in Economics, awarded to Philippe Aghion, Peter Howitt and Joel Mokyr, underscored how innovation-driven growth depends on a constant cycle of competition and renewal, what Joseph Schumpeter once called ‘creative destruction’. Their research showed that when dominant firms entrench their position, the economy’s capacity for innovation weakens.
That lesson is especially relevant in the AI era, where control over data and computing resources can act as a modern-day barrier to entry. The Nobel Committee’s recognition of this work serves as a reminder that prosperity in a digital economy depends as much on openness and contestability as on invention itself.
Recent reforms have tried to anticipate that. The Competition Commission of India now applies a “deal value” threshold that allows it to review acquisitions which may not be large in revenue but significant in strategic importance. Experts see it as a crucial move in keeping markets open to new entrants.
Still, experts caution that laws alone are not enough. As AI becomes embedded in every sector, from healthcare to logistics, ensuring access to data, compute and customers will determine whether innovation flourishes or fades.
Human factor
The AI story is being written at the intersection of innovation and regulation. For any country hoping to lead in AI, the task is to design systems that are not just powerful but fair, transparent and inclusive.
That means building algorithms that can be explained, audited and trusted — and making sure the benefits of AI do not stop at the gates of the biggest firms.
The next decade will test whether India can strike that balance. The answer, as Rai and Vishnoi suggest in different ways, lies in keeping human judgment at the centre of an increasingly automated world.
