About this site
As Amended holds Indian tax and corporate law in the regulator's own words, and records which notification changed which provision, and when.
Designed by Claude · Curated by Devansh Parikh
What it does
It collects Acts, Rules, Regulations and Master Directions from the authorities that issue them, keeps the text as published, and files every notification and circular against the provisions it affects. Then it links back: every screen carries the regulator’s own document, so nothing here has to be taken on trust.
Currently holding 186 instruments, 5,600 provisions and 8,053 notifications and circulars.
It runs by itself. Collection happens once a day without anyone reviewing what it brought back, which is a deliberate choice and the reason the next section exists.
Questions
Why are amendments not applied to the text?
Because applying them would mean a machine rewriting statutory wording, and then you would be reading the machine rather than the regulator. An amendment here is a reference: it tells you which notification changed this provision and when, and links you to it. The provision itself stays exactly as the regulator published it.
This is why the count of applied amendments is zero and will stay zero. It is the design, not a feature that has not been built yet. Follow the amendment link to read the change in the issuing authority’s own words.
How current is the text?
Collection runs daily at 07:00 IST. But the more useful answer is that the text is as current as the regulator’s own consolidation — if the income-tax department states a section as on a particular date, that date is shown on the provision, and that is what you are reading.
The “Last checked” stamp in the corner of every page says when the collector last looked at the regulators’ sites. Those are different questions and both are shown, because “as amended” means nothing without a date attached.
Can I rely on this for a filing or an opinion?
No — rely on the official document, which is linked from every screen. This site is built to get you to that document quickly and to tell you what has been issued against it. Treat it the way you would treat a good index: trustworthy enough to work from, never the thing you cite.
The terms say this more formally, and mean it.
Where does the text actually come from?
From the issuing authority, directly: the income-tax department, the Reserve Bank, SEBI, the Ministry of Corporate Affairs, the CBIC and India Code. Nothing is copied from a commercial publisher or from another website.
Each provision states whose rendering it is. Where the regulator publishes formatted text, that formatting is what you see — tables, provisos, indented clauses and footnote marks included. Where an instrument exists only as a PDF, the PDF is served and any text read out of it is marked as a finding aid rather than as the official rendering.
Something looks wrong — a section is empty, or holds the wrong text.
That happens, and the site keeps a public list of it rather than pretending otherwise. Provisions whose stored text does not look like a provision — a heading captured without its body, a row of a table, an amendment footnote — are flagged and listed on Text that needs a look.
The official document link on the provision is unaffected by any of this, and is the way through when the parsed text has let you down.
Is there a feed, or a way to get the data in bulk?
Yes to both, and both are preferable to crawling. An Atom feed carries every notification and circular as it arrives, and it can be narrowed to one regulator or one instrument with ?regulator= or ?instrument=. Outlook and Teams both read Atom, so it needs nothing installed.
There are also CSV exports of the documents and of the section map. Both are rate-limited.
Does it track me?
No. Reading this site needs no account, sets no cookie, loads no third-party script of any kind and shows no advertising. Page views are counted in aggregate at the moment they are written, so a record of an individual visit never exists — which is also why the site can report views and not visitors.
There is one deliberate exception, and it only happens if you ask for it. You can enter an email address to have the site remember the sections you follow, the ones you bookmark, the notes you write and the searches you keep. That holds personal data about you, which is why it is optional, never required to read anything, and deletable in one click — the delete button erases the account and everything in it at once, and the export button hands you the whole lot as a spreadsheet first.
Even then there is still no cookie: the account is kept in your browser’s own local storage and is never attached to an ordinary page request. And nothing is sent to an address until somebody opens a link sent to that address — ticking the box is not enough, deliberately, because a tick is made by whoever is holding the browser.
The full account is in the privacy notice, and the short answer on cookies is on the cookies page.
Who publishes it, and what does it cost?
Devansh Parikh, as an individual — not a firm, and not in any professional capacity. It is free, there is nothing to buy, and there is no intention to charge for it.
It runs entirely on free infrastructure, which is a design constraint rather than an accident: it is what keeps the site able to exist without needing to make money from the people reading it.
Recent changes
Material changes to the privacy notice or the terms are noted here.
- 15 October 2026 — the daily digest can now be emailed to readers who ask for it. Nothing is sent to an address until somebody opens a link sent to that address, and every message carries a one-click unsubscribe. Google is now listed as a processor on the privacy notice, because the service carrying a message necessarily handles the address it is going to. The weekly option was removed before it had ever sent anything; the digest is daily or nothing.
- 14 October 2026 — optional accounts added. Enter an email address and the site remembers your follows, bookmarks, notes and saved searches; it is never required to read anything. The privacy notice was rewritten to say exactly what is held, where it is held, and how to export or erase it, and the terms gained an Accounts section. The earlier wording — “no account, no cookie, no third-party script of any kind” — was true of the site that published it and became false on the first of those three the day accounts shipped; it has been corrected rather than quietly dropped. Still no cookies: the account lives in your browser’s local storage. A contact address for privacy questions and grievances is now published, which the previous wording admitted it lacked.
- 11 October 2026 — first publication of the privacy notice, cookies page and terms. Page-view counting introduced, aggregated and without any identifier.