When a Machine Starts Deciding Who Matters
Think about your last morning. You checked your phone, maybe applied for something online, maybe walked past a CCTV camera without even noticing it. Somewhere in that ordinary routine, a piece of software may have already made a decision about you, whether you get a loan, whether your face gets flagged at a metro gate, even which patient a hospital’s software nudges toward the next available bed. This isn’t science fiction anymore. It’s just Tuesday. Artificial intelligence has quietly slipped into courtrooms, banks, hospitals and welfare offices across India, and in doing so, it has walked straight into territory that Article 21 of our Constitution was built to guard: the right to life and personal liberty, which our courts have long read as including something harder to pin down, the right to live with dignity.
So here’s the question this piece tries to answer: what happens to that dignity when the decision-maker is no longer a person you can look in the eye, but a line of code you can’t even see? This isn’t an argument against AI. It’s an argument that we’ve already been handed the tools to deal with it, sitting right there in Maneka Gandhi v. Union of India[1] and Justice K.S. Puttaswamy v. Union of India[2]. We just haven’t turned those tools into working law yet, and the longer we wait, the wider the gap grows between what the Constitution promises and what actually happens to an ordinary citizen standing in front of a screen.
Dignity: The Quiet Word Hiding Inside Article 21
Read Article 21 on its own and it sounds almost plain: no person shall be deprived of life or personal liberty except according to procedure established by law.[3] For years, courts took that line at face value. If Parliament passed a law, however harsh, and followed the correct paperwork, that was that. Then came Maneka Gandhi[4], and everything shifted. The Court said a procedure isn’t good enough just because it exists on paper, it has to be fair, just and reasonable. That one sentence turned Article 21 from a technicality into something with teeth.
Dignity crept into this picture slowly, case by case, until it became less a specific right and more the reason all the other rights exist. By the time Puttaswamy[5] arrived in 2017, recognising privacy as part of Article 21, dignity had become the thread running through bodily autonomy, control over one’s own data, freedom from humiliating treatment, the whole basket. The judges were blunt about it: privacy protects your ability to be the author of your own story. Hold that thought, because it’s exactly what gets threatened when a machine starts writing that story for you.
Here’s the tricky part. Dignity isn’t a number you can measure or a box you can tick. It behaves more like a compass than a rulebook, it points you in the right direction without spelling out every step. Courts have leaned on it to strike down arbitrary state action, to protect prisoners and sex workers and manual scavengers, to insist that even the poorest person in the country keeps a core of respect that no policy can bargain away. AI is now testing that compass in terrain it was never built for, because the harm it causes is often invisible until it’s already happened to you.
AI Wears Two Faces, and Only One of Them Smiles
Let’s be fair to the technology first. AI has genuinely opened doors that used to stay shut. A diagnostic tool can screen for diabetic eye disease in a village clinic that has never had an eye specialist. A translation model can turn a government form into a language someone actually reads, instead of forcing them to pay a tout who “helps” for a fee. These aren’t small wins. They’re the kind of quiet, unglamorous progress that changes whether a person gets treated in time or gets left behind.
But flip the coin. The same pattern-spotting power that reads an eye scan can also reject your loan application, and here’s where it gets uncomfortable, without telling you why. A doctor who misses something on an X-ray can be questioned, dragged into a malpractice hearing if needed, made to explain the reasoning out loud. Ask an algorithm the same question and, often, nobody can answer, not even the engineers who built it. This is the famous “black box” problem, and it isn’t just a technical headache. Article 21’s demand for fair procedure has always implied a right to know why something happened to you, and a chance to push back.[6] A system that can’t explain itself simply can’t meet that bar, no matter how accurate it claims to be.
Picture a bank loan officer sitting across the table. Say no to the loan, and you can ask why, point out what he missed, maybe appeal to someone above him. Now replace him with a credit-scoring algorithm. The rejection lands in your inbox with no real explanation, maybe your pin code looked “risky” on paper, maybe a pattern in your spending resembled fraud statistically even though it wasn’t fraud at all. You’ve lost something real, not your life in the biological sense, but a shot at the kind of economic footing that makes a dignified life possible, and you lost it without anything close to the fair procedure Maneka Gandhi promised you.
[1]Maneka Gandhi v. Union of India, AIR 1978 SC 597.
[2]Justice K.S. Puttaswamy v. Union of India, (2017) 10 SCC 1.
[3]The Constitution of India, Art. 21.
[4]Supra note 1.
[5]Supra note 2.
[6]Supra note 1.