Gig Workers Got New Rights in 2026. Making Them Real Is Harder

Open any food or grocery app in an Indian city and a person on a two-wheeler is already weaving through traffic to bring your order. That person is part of a workforce NITI Aayog expects to swell from 7.7 million in 2020-21 to 23.5 million by 2029-30. For years these gig workers lived in a legal grey zone: not quite employees, not quite independent businesses, and so covered by almost nothing. In 2026, that finally began to change. The harder question is whether the new rules will mean anything on the road.

This is an analysis, not a verdict on any government. The aim is plain: lay out the genuine problems ordinary riders and drivers face, give fair credit for what has been built, and set out the specific reforms that labour economists and worker groups keep returning to.

What a rider actually deals with

Start with the lived reality, because policy that ignores it tends to fail. A widely cited field report titled Prisoners on Wheels found that around 80% of gig workers put in over 10 hours a day, nearly half get no weekly day off, and almost all report physical or mental strain from the grind. The Economic Survey has noted that around 40% of gig workers earn under ₹15,000 a month before they subtract fuel, phone data, and the cost of running and repairing their own vehicle.

Then there is the part that workers find hardest to fight: the algorithm. Order allocation, surge pay, ratings and penalties are decided by software that explains nothing. Two near-identical trips can pay differently, and no one tells you why. When an account is suddenly deactivated, there is often no manager to appeal to, only an automated chat window. In January 2025, the Indian Federation of App-Based Transport Workers told the Labour Ministry that workers simply cannot see how their earnings, ratings, task allotment and suspensions are decided.

Why the old labour rulebook never fit

India's labour protections were built for a factory-and-office world: a named employer, a fixed workplace, a monthly salary. Gig work breaks all three assumptions. A delivery partner may log into two apps in a morning, switch off at noon, and be classified as a "partner" rather than an employee on either.

That classification is not a technicality. It historically decided whether someone got provident fund, paid leave, accident cover or a path to complain about unfair treatment. Because platforms treated workers as independent contractors, the entire safety net that a salaried employee takes for granted simply did not apply. The result was a fast-growing, visible, essential workforce with close to zero formal cushioning.

What the government has actually built

Here the record deserves fair acknowledgement, because real scaffolding now exists. The Code on Social Security, 2020 came into force on 21 November 2025, and the four labour codes are slated for national rollout around 1 April 2026. For the first time, central law defines a gig worker and a platform worker as distinct categories that the state is responsible for, rather than leaving them in limbo.

The Code does several concrete things:

• It lets the government frame schemes for accident, health, maternity and old-age cover for gig and platform workers.
• It requires aggregators to contribute, with contributions pegged at a small share of turnover, into a dedicated Social Security Fund.
• It ties benefits to free registration on the e-Shram portal, which issues an Aadhaar-linked ID so that cover can travel with the worker across apps.
• From the 2026 rollout, a worker generally needs at least 90 days of platform work in a financial year to qualify for government-backed benefits.

States have moved too. Rajasthan became the first to pass a dedicated gig welfare law back in July 2023, built around a welfare cess on platform transactions, though it is still awaiting full operation. Karnataka went further in 2025, bringing in an ordinance and then a Bill, with a notification in February 2026 that mandates platforms to collect a welfare fee of 1% to 5% of the payout on each transaction for a workforce of reportedly more than 500,000 riders and drivers. Notably, Karnataka's law also writes in a worker's right to refuse a task without being punished for it, something the earlier Rajasthan framework did not spell out.

Taken together, this is a genuine shift from treating gig workers as nobody's responsibility to treating them as a recognised class with a funded welfare pool. That is progress worth stating plainly.

Where the gap between law and life remains

Recognition is not the same as protection. The Code enables schemes; it does not by itself guarantee a minimum wage, cap working hours, or force a platform to explain a deactivation. Several stress points stand out.

First, funding and registration. A welfare fund only helps if money flows in reliably and workers are actually enrolled. Patchy registration, confusion over who counts, and the 90-day threshold could leave the most casual, lowest-earning workers outside the net.

Second, algorithmic accountability. None of the headline measures yet require a platform to disclose how pay or ratings are set, or to justify cutting someone off. For a worker whose entire income depends on an app's good standing, an unexplained suspension is closer to a sacking than a glitch.

Third, portability and overlap. With central labour codes and separate state laws (each with its own cess and registration), platforms and workers face a tangle of rules that could mean double compliance in some places and gaps in others.

The reforms experts keep coming back to

The encouraging part is that the fixes are not mysterious. Across labour economists, worker collectives and policy researchers, a fairly consistent shortlist appears.

1. Make the algorithm legible. Require platforms to give workers a plain breakdown of how each payout, rating and penalty is calculated, and to flag major changes in advance. Transparency does not mean revealing trade secrets; it means a rider can understand their own paycheck.
2. Guarantee due process before deactivation. A clear, time-bound appeal heard by a human, with reasons given and a right to reply, would turn a sudden cut-off into a fair process. This is arguably the single change workers ask for most.
3. Set a fair earnings floor. Tie effective pay to working time, not just per-order rates, so that idle waiting and rising fuel costs do not quietly push real income below subsistence.
4. Build one portable benefits account. Let contributions from every platform a person works on flow into a single e-Shram-linked account, so cover follows the worker rather than resetting each time they switch apps.
5. Harmonise central and state rules. A common registration and a single, predictable cess structure would cut compliance friction and close the cracks workers currently fall through.
6. Add safety basics. Heat shelters, rest points, accident insurance that actually pays out quickly, and protection against unpaid waiting time address the daily physical reality of the job.

What to watch next

The 2026 rollout is the test. The legal architecture now exists, the welfare fund mechanism is defined, and at least two states have shown that a cess-funded model can be written into law. Whether that translates into a rider getting a clear explanation for a deactivation, an insurance payout after a road accident, or a pension contribution that survives a switch from one app to another, will be decided in the implementation details over the coming months.

The gig economy is not going away; convenience has made it part of urban life. The reasonable goal is not to break the model but to make it fair, so that the person bringing dinner to the door is building some security of their own while doing it. India has taken the first real steps. Finishing the job is mostly about turning paper rights into everyday practice.