Al: Who's Really in Charge?

                                                                      Professor R. S. Sengar and Er. Kartikeya

It's never just about who builds the model. It's about who gets to decide what data matters, what level of error is acceptable, and, crucially, who's left holding the bag when systems misfire. The 2026-27 Union Budget didn't cave to the AI (Artificial Intelligence) hype, nor did it sit on its hands. It walked a middle path, focused on building capacity, not chasing splashy headlines or moonshots. When I look at the government's Al approach, it falls into three buckets.

First, there is the heavy lifting, ramping up computing muscle, improving data infrastructure, and fueling the broader deep-tech space. Second, it's about people skilling, re-skilling, and yes, even training government officials in Al and ML (Machine Learning). (Quick fact: statistical officers made the training list in the Output-Outcome Monitoring Framework, the prized document that is tabled with the Budget. Who saw that coming?) Third, the Budget steers Al toward practical, testable domains such as agriculture, where results show up more clearly and faster.

What the Budget Says Out Loud

During her 1 February 2026 address, Finance Minister Nirmala Sitharaman introduced a standing committee to track how emerging tech, especially Al, is shaking up jobs, skills, and the entire education-to-employment chain. She also signaled intent to embed Al into school curricula and use it to match workers to jobs and training pipelines. One standout Initiative, Bharat-VISTAAR (Vision for Innovative Solutions and Technology for Agriculture Advancement and Reach) - A multilingual.

Al tool designed to fuse Agri Stack portals and ICAR (Indian Council of Agricultural Research) packages into tailored advisory systems for farmers. Another: under the Divyang Sahara Yojana, the government plans to scale assistive devices using Al and R&D (Research and Development), with distribution channels such as Assistive Technology Marts in tow.The real message: Skip the sizzle. Build systems that survive tenders, audits, and the chaos of last-mile delivery.

But That's Just the Start

Of course, it is hard to argue with this Budget. It does its job. But from 2027-28 onwards, we'll need clearer answers to some big-ticket questions:

Whose Al is it?

Who sets the terms?

Who wins from it?

And who's left paying when it fails?

Truth is, for early movers, Al isn't neutral technology. It's strategic weaponry. The US flexes through platforms. China scales. Europe regulates and exports its rulebook. The first movers write the defaults. Everyone else inherits the consequences.Late adopters don't just get technology, they inherit dependencies: on data standards, on vendor terms, on audit frameworks. Suddenly, they're negotiating on someone else's field, with someone else's referee. And 'keeping up' becomes less about tech, more about building strong institutions that can push back, question defaults, and rewrite rules where needed. But here's the snag: institutions move slowly. Markets don't. That gap is where silent dependencies creep in and stay.

India's Al Crossroads

Without clean data, solid connectivity, and a grievance redressal system that actually works, Al doesn't stand a chance. And right now, that infrastructure's still patchy across India.Sure, some argue that Al can fix the very problems that hold it back. But let's not forget: the same system that patches a leak can also quietly lock out the most vulnerable. One wrongly flagged file, one SMS rejection, and a welfare recipient might find himself out in the cold, with no clear way to appeal.

The promise of jobs is tangled in this mess too. Skill development sounds great on paper. But unless private sector hiring and public procurement catch up, skilling risks becoming a box-ticking exercise, well-intentioned but ultimately hollow. At least the Budget acknowledges this tension through its Education to Employment and Enterprise initiative. It's a nod to the gap between training and actual job creation.

Don't Forget the Ethics

Ethics isn't an afterthought. It needs to be baked into the system from day one consent protocols, privacy guards, bias audits, human oversight where it matters, and appeal mechanisms that actually function. Once Al systems go live at scale, retrofitting them is a bureaucratic nightmare. So yes, Budget 2026 plays it safe with its 'capacity now, scale later' bet. Probably a wise move. But what comes next must face the hard stuff: questions of power, accountability, fairness, and institutional backbone. Because in the end, a modernising state isn't judged by how fast it automates. It's judged by how well it protects its people in the process.

Writer: Professor R. S. Sengar, Director Training and Placement, Sardar Vallabhbhai Patel University of Agriculture and Technology, Modipuram, Meerut.