This page is informational and is not legal advice. It describes rules, orders and rulings as published or reported on the dates shown; read the primary text before relying on any of it.
Doorstep banking has had a rulebook since 2022. What changed on 4 December 2025 is narrower and more consequential: a single clause in RBI's new Branch Authorisation Directions for urban co-operative banks (UCBs) decides, for the first time by deposit size, whether a bank may use a commission agent for collection at all. We want to set out exactly what that clause says, what follows from it arithmetically, and — just as importantly — what it does not settle.
The clause
Clause 74 of the Directions reads: "Tier 3 and 4 UCBs may offer doorstep banking services to customers either directly through its own employees or through agents." It then provides that "Tier 1 and 2 UCBs shall offer doorstep banking services only through their permanent employees." The annex to the Directions repeats the point as a mode-of-delivery restriction: agents are "permitted only for Tier 3 and 4 UCBs." Pigmy collection sits inside this framework because the first permitted doorstep service the annex lists is "pick up of cash against receipt."
RBI's tiers are set by deposit size, not by any other measure of a bank's activity:
| Tier | Deposits |
|---|---|
| Tier 1 | All unit and salary earners' banks regardless of size, plus other banks up to ₹100 crore |
| Tier 2 | More than ₹100 crore and up to ₹1,000 crore |
| Tier 3 | More than ₹1,000 crore and up to ₹10,000 crore |
| Tier 4 | More than ₹10,000 crore |
Read plainly, a UCB with deposits under ₹1,000 crore may run doorstep collection only through permanent employees on its payroll from December 2025 onward. The commission agent — the delivery mechanism pigmy collection has used since Syndicate Bank launched the scheme in 1928 — is not available to that bank for this purpose.
Who this actually touches
How many urban co-operative banks sit below the ₹1,000 crore deposit line decides how large a change this is, and we have not verified a current count. A bank that has run pigmy collection on commission agents for decades now has to decide whether to convert those collectors to payroll, restructure the scheme, or stop offering doorstep collection altogether — and RBI's own tiered framework gives no transition period specific to this clause.
Set against that, a large amount of doorstep collection sits entirely outside this rule. Clause 74 governs UCBs; it says nothing about credit co-operative societies, patpedhis, Souharda co-operatives, multi-state credit societies or Nidhi companies, because none of them are UCBs and none of them answer to RBI. In a state like Maharashtra, where roughly two lakh pigmy agents serve close to a crore families across some 16,000 credit co-operatives, the clause changes nothing — that entire agent base sits with institutions the state Registrar regulates, not RBI. The rule reaches deep into the UCB segment and does not touch the segment where most agents, by the only published count we have, actually work.
Tier 3 and 4 UCBs are also unaffected in substance: they may keep using agents, subject to the same Board-approved policy, published charges and cash limits that have applied to doorstep banking since the 2022 circular that first set the framework out.
What we want to be careful about
The text of clause 74 is not ambiguous about who may use an agent. What is genuinely open is how it applies to a pigmy scheme that predates the doorstep banking framework itself. Many UCBs have run a pigmy collection round for decades, under whatever internal scheme or byelaw governed it at the time. Whether a bank's supervisors treat that legacy scheme as the "doorstep banking service" the Directions define in Chapter V, or as something outside that chapter's reach, is a question we have not found addressed in any published guidance. We are not asserting an answer either way, and anyone telling you with confidence how every legacy scheme will be treated is going beyond what the text settles.
A second open question is scale. We found no current, published, tier-wise count of how many urban co-operative banks actually sit in each tier. It is reasonable to say the below-₹1,000-crore band covers most of the sector by number of institutions, because that is how RBI's own tier bands are constructed, but we cannot put a precise figure on how many banks, or how many existing agents, this clause reaches. Treat any specific count you see elsewhere as an estimate unless it cites a primary source.
A third question sits at the boundary with a separate development. In April 2026 the Karnataka High Court held, in a matter involving a Regional Rural Bank, that the bank's pigmy agents were employees rather than independent business facilitators, for GST purposes. That is a different institution type, a different regulator and a different question of law from clause 74 — RRBs are not UCBs, and the case is about tax, not about who a bank may appoint — but the two developments point in an interesting direction together: one framework is narrowing where commission agents may be used at all, while another has found, on one set of facts, that an agent relationship already looked like employment. We cover that ruling and its own limits separately, and we would caution against reading either development as resolving the other.
What this means in practice
If you run a UCB below the ₹1,000 crore threshold, the practical starting point is not the agent-selection policy — it is confirming your tier, then deciding how your existing collectors are engaged going forward. We set out the board-approval, notification and policy steps in appointing pigmy agents. If you are a depositor, the tier rule does not change your receipt, your crediting timeline or your deposit insurance position; those depend on the institution type and the doorstep rules covered in the rules behind pigmy collection, not on which staffing model the bank uses to deliver the visit.
Nothing here is financial or legal advice. It is a reading of a published clause, held carefully at the edge of what it actually says.