Who pays when an Ola or Uber driver causes an accident

An accident caused by a ride-hailing driver — Ola, Uber, Rapido, inDrive — raises a vicarious-liability question that Indian tort and motor-vehicle law are still calibrating. The driver is the primary tortfeasor at common law. The registered owner of the vehicle is vicariously liable under the master–servant doctrine recognised in Sitaram Motilal Kalal v Santanuprasad Jaishankar Bhatt, AIR 1966 SC 1697 / (1966) 3 SCR 527, State Bank of India v Shyama Devi, AIR 1978 SC 1263 and Pushpabai Purshottam Udeshi v Ranjit Ginning and Pressing Co (P) Ltd, AIR 1977 SC 1735. The third-party insurer is bound to satisfy any award under Section 149 of the Motor Vehicles Act, 1988 read with the framework laid down in National Insurance Co Ltd v Swaran Singh, (2004) 3 SCC 297. The harder, unsettled question is whether the aggregator platform is itself vicariously liable. The 2019 amendments to the Motor Vehicles Act inserted the statutory definition of "aggregator" in Section 2(1A) and obliged the State to license aggregators under Section 93; the Motor Vehicle Aggregator Guidelines, 2020 (revised 2025) read into the licence obligations that resemble principal-style duties — driver vetting, fare regulation, complaint resolution, insurance arrangements. This guide walks through the layered liability map and the aggregator-as-principal question that the High Courts are presently working through.

A ride-hailing trip is, doctrinally, a tangle of relationships. The passenger has booked through an aggregator's mobile application; the trip is performed by a driver who is registered with the platform and (in most cases) owns the vehicle in his own name; the vehicle is insured under a Section 147 Motor Vehicles Act policy taken out by the driver-owner. When the trip ends in a road accident, the question "who pays" cuts across at least four candidates — the driver, the registered owner (often the same person), the third-party insurer of the vehicle, and the aggregator platform. The first three are settled. The classical Indian framework for vicarious liability — laid down in Sitaram Motilal Kalal v Santanuprasad Jaishankar Bhatt, AIR 1966 SC 1697, refined in State Bank of India v Shyama Devi, AIR 1978 SC 1263 and Pushpabai Purshottam Udeshi v Ranjit Ginning and Pressing Co (P) Ltd, AIR 1977 SC 1735 — supplies the master–servant test for the driver–owner relationship. The Supreme Court's framework in National Insurance Co Ltd v Swaran Singh, (2004) 3 SCC 297 supplies the insurer's duties. The harder question — whether the aggregator is also vicariously liable, on what footing, and to what extent — is being worked out under the 2019 amendments to the Motor Vehicles Act, which introduced the definition of "aggregator" in Section 2(1A), made aggregator licensing mandatory under Section 93, and obliged the Ministry of Road Transport and Highways to issue the Motor Vehicle Aggregator Guidelines, 2020 (revised in 2025). This article works through each layer.

The classical framework — Sitaram and Shyama Devi on master–servant vicarious liability

The Indian doctrine of vicarious liability for a road accident rests on the master–servant relationship: a master is liable for the wrongful acts of a servant committed in the course of employment. The two essentials, laid down in a long line of cases, are that the tortfeasor must be a "servant" of the principal, and the wrongful act must be done within the scope of the servant's employment. The Supreme Court applied the framework to a vehicular accident in Sitaram Motilal Kalal v Santanuprasad Jaishankar Bhatt, AIR 1966 SC 1697 / (1966) 3 SCR 527. Sitaram, the registered owner of a car used as a taxi, had entrusted the vehicle to one Mohammad Yakub as the driver. Yakub, in turn, allowed the cleaner to take the taxi for a driving test; an accident followed. The Supreme Court held that the cleaner was not the owner's servant — there was nothing to show that Sitaram had either permitted the cleaner to drive or had authorised Yakub to delegate the driving authority to a stranger. The cleaner was, at the time of the accident, not doing the master's work; nor was Yakub, in lending the vehicle, acting in the master's business. The owner was held not liable.

The complement to Sitaram is State Bank of India v Shyama Devi, AIR 1978 SC 1263. A customer of the bank gave cash and cheques to a bank employee — in his personal capacity as the customer's husband's friend — for being deposited into the customer's account; the employee misappropriated the funds. The Supreme Court held that when the employee committed the fraud he was not acting in the scope of the bank's employment but in his private capacity as the depositor's friend; the bank was therefore not vicariously liable. The case is the canonical Indian articulation of the principle that a tort committed by an employee outside the scope of employment, even if the opportunity for the tort arose from the employment, does not fasten liability on the master.

The "course of employment" test was generously interpreted in Pushpabai Purshottam Udeshi v Ranjit Ginning and Pressing Co (P) Ltd, AIR 1977 SC 1735. The manager of the company, while driving a company car on the company's business, gave a lift to another employee; an accident caused by the manager's negligence killed the lift-taker. The Supreme Court held that the act of giving a lift to a fellow employee was within the manager's ostensible authority and within the course of employment; the company was vicariously liable.

The framework was further refined in Rajasthan State Road Transport Corporation v Kailash Nath Kothari, (1997) 7 SCC 481, where the RSRTC had hired a truck along with its driver from a private owner. The Supreme Court held that the corporation, having effective control over the driver during the period of hire, was vicariously liable for the driver's negligence even though the driver remained on the original owner's pay-roll. The "effective control" test became, after Kailash Nath Kothari, the dominant Indian test for vicarious liability in transferred-employment situations — superseding the older "pay-roll" test.

Applying the classical framework to a ride-hailing driver — three settled answers

Applied to a ride-hailing accident, the classical framework supplies three settled answers and one open question.

The driver is the primary tortfeasor. No special doctrinal complication arises here. The driver who caused the accident through rash, negligent or impaired driving is personally liable in tort for the resulting injury or death. This much is uncontroversial.

The registered owner is vicariously liable as the master. Where the driver is also the registered owner — the common pattern for Ola, Uber and Rapido vehicles where the driver-partner has bought the car under a personal loan or hire-purchase arrangement — the question collapses into the personal liability of the driver-owner. Where the registered owner and the driver are different persons — a "fleet partner" who owns several vehicles operated by hired drivers — the registered owner is the master, the driver is the servant, and vicarious liability arises in the ordinary way under Sitaram and Pushpabai. The 2019 amendment to Section 2(30) of the Motor Vehicles Act, 1988 kept the broad statutory definition of "owner" intact — "a person in whose name a motor vehicle stands registered", and where the vehicle is the subject of a hire-purchase, lease or hypothecation agreement, the person in possession under the agreement.

The third-party insurer is bound to satisfy the award. The compulsory third-party policy under Sections 146 and 147 of the Motor Vehicles Act, 1988 is in force for every vehicle on the road, including a ride-hailing vehicle. The insurer's defences are narrowly drawn under Section 149(2) — they are the breach-of-policy defences listed in clauses (a) to (c) of the sub-section. The Supreme Court's framework in National Insurance Co Ltd v Swaran Singh, (2004) 3 SCC 297 made the third-party-protection principle dominant: the insurer must first pay the third-party victim and may then pursue its pay-and-recover remedy against the owner where a defence is made out. The Supreme Court's larger-bench ruling in Mukund Dewangan v Oriental Insurance Co Ltd, (2017) 14 SCC 663 settled the related question whether a holder of a light-motor-vehicle (LMV) licence may lawfully drive a transport vehicle below 7,500 kg in unladen weight without a separate transport-vehicle endorsement — the bench held in the affirmative. The ruling matters for ride-hailing because many driver-partners drive their personal cars used commercially on LMV licences; the Mukund Dewangan framework forecloses the insurer's defence on the licence-mismatch ground in the standard case.

The aggregator's vicarious liability — the open question. This is where the classical framework strains. The aggregator's relationship with the driver is structured, on paper, as that of a "technology service provider" with an "independent contractor" or "driver-partner". The aggregator's terms-of-service almost always disclaim any employer–employee relationship; the driver is described as an independent businessperson using the aggregator's matchmaking software in return for a commission. Applying the classical "course of employment" test, the driver would not be the aggregator's servant; the aggregator would not be vicariously liable. But the post-2019 statutory framework complicates the analysis.

The 2019 statutory pivot — Section 2(1A), Section 93 and the 2020/2025 Aggregator Guidelines

The Motor Vehicles (Amendment) Act, 2019 introduced three statutory changes that bear on aggregator liability.

Section 2(1A) — definition of "aggregator". The 2019 amendment inserted the statutory definition: "aggregator" means a digital intermediary or marketplace for a passenger to connect with a driver for the purpose of transportation. The definition is intermediary-flavoured — it borrows the "marketplace" vocabulary familiar from e-commerce — but it also recognises that the aggregator connects the passenger with the driver, not merely the driver with the vehicle.

Section 93 — licensing of aggregators. The 2019 amendment substituted Section 93 to require every aggregator to hold a licence issued by the State Government in accordance with such conditions as may be prescribed. The licence requirement is mandatory; an unlicensed aggregator commits an offence. The State Governments have, in compliance, framed state-level aggregator policies (Karnataka, Maharashtra, Delhi, Tamil Nadu, Telangana have each issued state guidelines or rules).

The Motor Vehicle Aggregator Guidelines, 2020 (revised 2025). The Ministry of Road Transport and Highways issued the Guidelines under Section 93 in November 2020; a revised version was issued in 2025 to address commission caps, surge-pricing limits, driver insurance, women-safety obligations and grievance-redress mechanisms. The Guidelines impose obligations on the aggregator that resemble principal-style duties: (a) the aggregator is to verify the driver's licence, antecedents and police record before on-boarding; (b) the aggregator is to ensure the vehicle is properly insured (including, in the 2025 version, a recommended insurance cover for the driver as a "platform worker"); (c) the aggregator is to operate a grievance-redress mechanism with defined response times; (d) the aggregator is to maintain trip records and share them with the State Government on demand; (e) the aggregator is to comply with fare caps and surge-pricing limits set by the State Government.

The doctrinal question that follows from this framework is whether the cumulative effect of the statutory licensing, the verification obligations, the trip-record duties and the grievance-redress duties converts the aggregator from a pure technology intermediary into a principal — at least for the limited purpose of vicarious liability for the driver's on-platform torts. The answer is not yet authoritatively settled by the Supreme Court.

The aggregator-as-principal argument — what supports it

Three lines of argument run in favour of aggregator vicarious liability.

The "effective control" test from Kailash Nath Kothari. Although the aggregator's terms-of-service describe the driver as an independent contractor, the on-platform reality looks closer to effective control. The aggregator: assigns the trip; sets the fare; restricts driver discretion to refuse fares (a high refusal rate triggers de-platforming under most aggregator policies); rates and disciplines the driver; sets vehicle and behavioural standards; collects the fare and remits the driver's share; suspends or terminates the driver-partner relationship at its discretion. The Kailash Nath Kothari test — effective control rather than pay-roll — would, if applied generously, place the driver within the aggregator's "course of employment" for on-platform trips.

The Code on Social Security, 2020. The Code recognises "gig worker" (Section 2(35)) and "platform worker" (Section 2(60)) as statutory categories distinct from "employee" but entitled to defined social-security benefits (life and disability insurance, health and maternity benefits, old-age protection). The Code's recognition that an aggregator owes statutory obligations to its platform workers — even if those workers are not "employees" in the classical sense — is an instance of the legislature looking past the technology-intermediary framing.

The agency framework in Pushpabai. The Supreme Court in Pushpabai applied a generous "ostensible authority" test — the manager-driver was acting within the company's ostensible authority when he gave a lift to a fellow employee. The aggregator-driver relationship, mediated through a continuously-running mobile application, displays features of ostensible authority on which a court applying Pushpabai could fasten — the trip is displayed in the application as an aggregator trip, the passenger is in a vehicle bearing aggregator livery (where the platform requires it), the fare is collected through the application, and the receipt is issued in the aggregator's name. The applied-Indian-agency framework — Sections 182 to 238 of the Indian Contract Act, 1872 — has the conceptual tools (ostensible authority, agency by holding-out, the doctrine of estoppel) that could support a finding of aggregator vicarious liability.

The aggregator-as-intermediary argument — what supports it

The countervailing argument rests on three legs.

The Information Technology Act intermediary safe-harbour. Section 79 of the Information Technology Act, 2000 confers a conditional safe-harbour on an "intermediary" — defined to include any person who, on behalf of another person, receives, stores, transmits or provides a service in respect of an electronic record. The aggregator's terms-of-service rely on the safe-harbour, characterising the platform as a Section 79 intermediary. The doctrinal counter is that the Section 79 safe-harbour is conditional on the intermediary observing "due diligence" — a duty that the post-2019 motor-vehicle framework arguably gives content to in the form of the MV Aggregator Guidelines obligations. An intermediary that fails the due-diligence duty loses the safe-harbour.

The independent-contractor framing. The classical Indian law of vicarious liability draws a sharp line between a "servant" (for whose torts the master is liable) and an "independent contractor" (for whose torts the principal is generally not liable, save the exceptions discussed in the textbooks — strict-liability operations, dangerous-on-the-highway works, nuisance, and statutorily-imposed common-law duties). The aggregator's contractual framing pushes the driver into the independent-contractor category; an unmodified application of the classical line would foreclose vicarious liability.

The Section 93 licence as a regulatory, not a private-law, framework. The aggregator's defenders argue that the Section 93 licence and the 2020/2025 Guidelines are regulatory instruments — they create public-law obligations enforceable by the licensing authority through licence suspension or revocation, not private-law duties enforceable by the accident victim through a tort claim. The argument is plausible but not unanswerable: regulatory obligations have routinely been read by Indian courts to give content to common-law duties (the entire body of negligence law applied to statutorily-licensed activities — from the Bhopal cases to the road-construction cases — proceeds on this premise).

Parallel routes — Consumer Protection Act, 2019 and the deficiency-of-service claim

Independently of the vicarious-liability question, the passenger has a deficiency-of-service route under the Consumer Protection Act, 2019. The Act, in Section 2(7), defines a "consumer" to include a person who hires any service for consideration; the ride-hailing passenger is a consumer of the aggregator's service. The 2019 Act, in Section 2(42), defines "service" broadly enough to cover ride-hailing. The 2019 Act, in Section 2(47), defines "unfair trade practice" widely and, in Section 94 (the e-commerce provisions), authorises the Central Government to make rules for the protection of consumers in e-commerce — under which the Consumer Protection (E-Commerce) Rules, 2020 were notified. The Rules expressly cover "marketplace e-commerce entities" and impose duties of disclosure, grievance-redress and consumer-protection on the platform.

The Supreme Court in Skill Lotto Solutions Pvt Ltd v Union of India, (2021) 15 SCC 667 affirmed the broad reach of the consumer-protection regime against intermediary-style entities — although the case did not involve transport aggregation, the bench's framework on intermediary liability has been read into the aggregator context by the District and State Consumer Commissions. A passenger injured in a ride-hailing accident can therefore frame a parallel deficiency-of-service complaint before the Consumer Commission, naming the aggregator as opposite party. The relief is not the same as a tort-law compensation award — the Consumer Commission's pecuniary jurisdiction (District up to Rs 1 crore under Section 35; State up to Rs 10 crore under Section 47; National above that under Section 58) and procedural framework differ from the MACT framework — but the route is available and is being increasingly invoked.

The MACT claim — the practical path for the injured passenger or third party

The practical path that the injured passenger or third-party victim of an aggregator-vehicle accident takes is the Motor Accident Claims Tribunal under Section 166 of the Motor Vehicles Act, 1988. The claim is filed against — at minimum — the driver, the registered owner and the insurer; the aggregator is increasingly being added as a co-respondent. The Tribunal's pecuniary jurisdiction is unlimited (the heads-of-claim framework discussed in the companion article on grievous-injury compensation applies). The third-party insurer's duty to satisfy the award is governed by Sections 147 to 149 read with the Swaran Singh framework.

Where the aggregator is added as a co-respondent and the Tribunal finds vicarious liability, the award runs jointly and severally against the driver, the owner, the insurer (to the extent of policy cover) and the aggregator. Several High Courts have, in unreported and trial-court orders, applied the Pushpabai ostensible-authority framework and the Kailash Nath Kothari effective-control framework to fasten partial vicarious liability on the aggregator; the Supreme Court has not yet authoritatively spoken on the point. The recent jurisprudential trend on related questions — the recognition of platform workers in Indian Federation of App-Based Transport Workers v Union of India (pending writ proceedings before the Supreme Court on the social-security framework) and the High Court rulings on aggregator-regulation matters — suggests that the doctrinal direction of travel is toward expanded aggregator liability.

What to watch for — pleading, evidence and the live questions

Three operational issues recur in aggregator-accident litigation.

Pleading the aggregator as co-respondent with specific facts. A claim that names the aggregator generically — "Ola/Uber is liable as the platform" — is liable to be dismissed at the threshold. The pleading must allege specific facts: that the trip was an on-platform trip (with the booking-ID), that the aggregator on-boarded the driver under the Section 93 licence framework, that the aggregator failed to discharge a specific obligation under the 2020/2025 Guidelines (driver vetting, insurance arrangement, vehicle-fitness verification), and that the failure was causally connected to the accident. Documentary anchors are the booking-ID, the trip-history record (obtainable from the application), the aggregator's communications, and the insurer's policy schedule.

The on-platform versus off-platform distinction. The vicarious-liability argument runs only for on-platform trips. A driver who has logged out of the application and is driving for personal reasons when the accident occurs is outside any plausible "course of employment" with the aggregator. The application's log-record is the determinative evidence; the aggregator's compliance with the trip-record duty under the 2020/2025 Guidelines makes the log obtainable. A claim that fails to establish on-platform status faces the Sitaram outcome.

The interplay with the insurer's pay-and-recover defence. Where the insurer pays the third-party victim and then seeks to recover from the owner under the Swaran Singh framework, a question arises whether the insurer's recovery rights run against the aggregator as well. The point is unresolved; the safer pleading is to seek a direct finding of joint-and-several liability that includes the aggregator, so that the insurer's recovery (if any) does not unwind the aggregator's contribution.

Outcome — the layered claim and the doctrinal frontier

The architecture of liability for a ride-hailing accident is layered. The driver is personally liable in tort. The registered owner is vicariously liable under the master–servant doctrine in Sitaram, Pushpabai and Kailash Nath Kothari. The third-party insurer is bound to satisfy the award under Sections 146 to 149 of the Motor Vehicles Act, 1988, read with the Swaran Singh framework, and may pursue its pay-and-recover remedy in the limited circumstances permitted by Section 149(2). The aggregator's liability is the live frontier — the post-2019 statutory framework (Section 2(1A) defining the aggregator, Section 93 licensing aggregators, the 2020 Aggregator Guidelines as revised in 2025) imposes obligations that look principal-style; the courts are presently calibrating whether and when those obligations translate into private-law vicarious liability for on-platform driver torts.

The doctrinal trajectory is unsettled but coherent. The Supreme Court's classical framework — Sitaram (1966), Shyama Devi (1978) and Pushpabai (1977) — supplies the master–servant and ostensible-authority tests. The Kailash Nath Kothari (1997) refinement — effective control rather than pay-roll — supplies the test that maps most naturally onto the aggregator relationship. The Code on Social Security, 2020 statutorily recognises the platform worker. The Section 93 licensing regime and the 2020/2025 Aggregator Guidelines impose principal-style obligations on the aggregator. The Consumer Protection Act, 2019 supplies a parallel deficiency-of-service route. The remaining questions — the precise reach of the aggregator's vicarious liability, the interplay with the Section 79 IT Act intermediary safe-harbour, and the contribution-and-indemnity framework between the aggregator and its on-platform insurance arrangements — are working themselves out at the High Court and tribunal level. Until the Supreme Court speaks, the operating manual is the classical framework as informed by the post-2019 statutory framework and the 2020/2025 Guidelines.

The practical lesson for the injured passenger or third-party victim is to plead the layered claim in full — the driver, the registered owner, the insurer and the aggregator — and to support each leg of the claim with specific evidence. The MACT route under Section 166 is the primary remedy; the Consumer Commission route under the Consumer Protection Act, 2019 is a parallel option that can be invoked for deficiency-of-service complaints against the aggregator. The architecture is doctrinally complicated but procedurally navigable.