A skateboard is not a means of transport pretending to be one. It's closer to handwriting, a way of putting a person's particular rhythm onto a shared surface, visible to anyone walking past. The line a skater draws through a plaza, up a ledge, along a rail, is a kind of authorship, temporary and physical rather than written, but authorship all the same. Cities have never quite known what to do with that. Wheels on concrete get filed under noise complaint or property damage long before anyone considers what's actually being expressed, and the surfaces that could hold it, empty lots, underused plazas, wide unused stretches of pavement, stay closed off by default rather than by any real decision that they should be. That default has mostly gone unquestioned because nobody's had a reason powerful enough to force the question, until the Supreme Court gave one, without meaning to. On June 19, the Court held that the right to walk on a safe, demarcated footpath is a fundamental right, arising from Article 19(1)(d) read against Article 21. The case reached that conclusion through a five year old boy, killed by a tanker on a road with no footpath, no crossing, nowhere for a person on foot to exist safely. Civic bodies, the Court said, now carry an enforceable duty to build and maintain that infrastructure, not a discretionary courtesy but an obligation with teeth. What makes the judgment worth sitting with, past the tragedy that produced it, is where the Court chose to locate the right. It didn't rest the reasoning purely on 19(1)(d)'s guarantee of movement. It read that clause alongside 19(1)(a), (b), and (c), speech, assembly, association, treating the ability to occupy public space as the physical condition those freedoms actually depend on. A footpath, in that framing, isn't only a route from one place to another. It's the ground a person needs under them before any of the other freedoms can be exercised at all. Skateboarding makes the same claim on the same ground and has never once been read that way. It asks for a surface, clear of traffic, wide enough to be used without harming anyone, exactly what the judgment now compels cities to provide for pedestrians. But the request has always been treated as recreation, a hobby to be tolerated somewhere out of sight, rather than as a form of movement carrying its own expressive weight, its own claim to public space on the same constitutional footing the Court just extended to walking. The difference isn't in the nature of the act. It's in the fact that walking eventually produced a body the state had to answer for, and skateboarding, so far, has mostly produced noise complaints and shrugged-off municipal notices instead. That's the idea worth holding onto here, separate from whatever the judgment technically covers. A city's willingness to design for a use of public space tends to track how much harm that use has caused, not how much freedom it actually represents. Skateboarding sits in the gap between those two measures, low on documented harm, high on the same claim to expression and movement the Court just spent several pages defending. Nobody has to be struck by a tanker for that claim to matter. It's already there, in every skater finding a ledge nobody built for them, writing something briefly onto a surface the city never intended to hold it.

Read the coverage and you'd think the Delhi High Court just declared AI training legal in India. It didn't. Justice Amit Bansal refused ANI an interim injunction because ANI couldn't prove ChatGPT had memorised or spat back its reporting, not because OpenAI walked in clean. Fair dealing gave the ruling its legal skin. The actual bone underneath is duller and more consequential, an argument about who has to bring proof, and proof of what. That's the part getting skipped past. The order quietly moves the weight of an entire category of dispute onto the party least equipped to carry it. A newsroom, a musician, an illustrator, anyone with a suspicion that their work fed a model somewhere, no longer gets to start from suspicion. They need an output that unmistakably wears their own signature, produced by a system built specifically not to show its working. Ask any of them how they're supposed to get that, and the honest answer is, mostly, they can't. The fair use finding will get cited for years, but it was always the easier question. Courts have been deciding what counts as fair dealing since long before anyone typed a prompt. The harder question, and the one this order actually answers, is procedural, not doctrinal, who walks into court holding the burden while the underlying technology stays a black box to everyone standing outside it. That answer just tilted, and it tilted before most people noticed there was a question being asked. Meanwhile ANI's other claim sits completely untouched, that the model didn't just train on its work but invented stories wholesale and stapled ANI's name to them. That has nothing to do with training data. It's what happens when a system fabricates something and dresses it in a real identity. Nobody's decided yet who has to prove what there, whether the burden falls on the name that got borrowed or the machine that borrowed it. Given where this order landed, it isn't hard to guess which way that one leans too. Which is the actual story here, buried under a headline about fair use. This wasn't a ruling about artificial intelligence. It was a ruling about who gets to stay silent and who has to speak first, dressed up in the language of copyright because copyright happened to be the case on the docket. The next dispute might be about a voice, a face, a melody, a sentence someone swears they wrote first. The technology underneath will keep changing. What this order set is the posture everyone starts from, proof-first or proof-later, and that posture, once it settles into precedent, tends to outlive the case that built it.

Every other art form gets to outlive its usefulness. A painting can hang unseen in a storeroom for decades and still be a painting. A manuscript can go unread and remain, legally and materially, exactly what its author made. Architecture doesn't get that grace. A building is art that has to keep working, keep housing something, keep justifying the land under it, and the moment it stops, someone with a legitimate claim to that land can simply have it removed. No other art form is quite so conditional on its own continued utility. Indian copyright law tried to account for this asymmetry through Section 57, the moral rights provision, which gives an author the right of paternity and the right of integrity, the right to be identified with a work and the right to object to its distortion or mutilation in ways that harm their honour or reputation. It's a good instrument, built for exactly the kind of harm architecture tends to suffer, unauthorised alteration, a facade stripped for renovation, a structure changed beyond recognition while still technically standing. What it was never built for, and what it turned out not to reach, is outright destruction. That gap became fact in Raj Rewal v. Union of India. Rewal had designed the Hall of Nations at Pragati Maidan, the first large span concrete structure of its kind in the country and, by most accounts, one of the defining works of Indian modernist architecture. In 2017, ITPO demolished it to redevelop the site. Rewal sued under Section 57, arguing the demolition was itself an assault on his reputation as the work's author, not a modification of the Hall but its erasure. The Delhi High Court disagreed. It held that moral rights protect a work from distortion while it exists, not from an owner's decision to end its existence entirely, and leaned on Section 52(1)(x), which permits a building to be reconstructed from its original plans without infringing anyone's copyright. If demolition itself were barred, the Court reasoned, that provision would have nothing left to permit. There's a strange logic sitting underneath that reasoning, worth pulling apart on its own terms. The Court effectively said a demolished building can always be rebuilt from its drawings, so the drawings are what survive, and the drawings are what the law was always protecting. But a reconstruction is not the thing itself, any more than a forgery made from a photograph is the original painting. The Hall of Nations was significant because it stood where it stood, built when it was built, aged the way concrete ages, carrying four decades of a country's history inside its own structure. None of that transfers to a replica built later from a technical drawing. The law found a way to call the object replaceable by treating the idea of it as the only part that mattered, which is a real philosophical move, and also a fairly convenient one for whoever wanted the land back. What the case actually exposes, once the doctrine is set aside, is a choice this country hasn't made yet, whether architecture is closer to infrastructure, disposable once its function ends, or closer to art, worth preserving regardless of function. Most legal systems that take moral rights seriously eventually have to answer that question directly, sometimes through heritage law, sometimes through cultural property statutes that sit outside copyright altogether. India has largely let property rights answer it by default, letting ownership of the land settle a question that ownership was never really equipped to settle. An architect's authorship doesn't disappear the day a building comes down. It just stops having anywhere to live. That's the idea worth holding past this one case. A society's relationship to its own built art shows up less in what it protects on paper and more in what it's willing to let disappear without a fight. Every demolished structure of real artistic weight is a small referendum on that question, decided quietly, usually by whoever holds the title, rarely by anyone weighing what the work itself was worth keeping. Preservation, in that sense, was never really a legal problem waiting for the right provision. It's a cultural one, about whether a country is willing to treat the buildings it stands inside as something closer to authorship than to real estate, long before any court has to be asked.