On 24 March 2023, a State Bank of India employee finished his usual train ride from Navi Mumbai to Kurla and took a shared autorickshaw for the last leg to his branch at the Bandra Kurla Complex, as he had for fourteen years. The driver seated three passengers in the back, as shared autos do. On the account recorded in the judgment, which comes from the petition since the complainant did not appear before the High Court, what happened next between the employee and a female co-passenger, a chef employed elsewhere, is disputed. He says the cramped seating on a bad road caused his hand to brush against her bag. She accused him of deliberately touching her inappropriately, used pepper spray on him, and called the police; he was taken to the BKC police station and a case was registered under Section 354A of the Indian Penal Code. She also complained to her own employer's Internal Committee, which forwarded the complaint to the bank, and the bank referred it to its own Internal Committee. By an order of 29 August 2023, that committee found him guilty of sexual harassment and recommended action under the service rules. What actually happened in that autorickshaw is not the subject of this piece, and nothing said here depends on which account is true.

The Bombay High Court has now set that order aside. Not because it examined whether the harassment happened, but because it held the Internal Committee had no jurisdiction to examine anything at all. In Siddhesh Pradeep Satpute v. State Bank of India & Ors., a division bench of Justices Suman Shyam and Firdosh P. Pooniwalla ruled that a shared autorickshaw hailed on the street is not a “workplace” under Section 2(o) of the Sexual Harassment of Women at Workplace (Prevention, Prohibition and Redressal) Act, 2013, unless the employer provided the transport (2026 LiveLaw (Bom) 290; neutral citation 2026:BHC-OS:13187-DB). Jurisdiction under the Act turns entirely on whether the incident occurred at a workplace. This one, the Court held, did not, so the Committee's order could not stand. The Court was careful to say it had decided nothing on the merits, leaving open whether the underlying conduct could still be pursued in an appropriate proceeding. But the jurisdictional question it did decide is one every employer, HR officer and Internal Committee in urban India now has to reckon with, because the reasoning sits awkwardly against how most people in Mumbai, Bengaluru or Delhi actually get to work in 2026.

What the Statute Says, and What the Court Read Into It

Section 2(o)(v) of the PoSH Act defines “workplace” to include any place visited by the employee arising out of or during the course of employment, including transportation provided by the employer for undertaking such journey. The petitioner argued that this transportation clause is the operative one: since the shared auto was public transport, not employer-provided transport, the incident did not happen at a workplace, and a committee that has no workplace has no jurisdiction. The bank defended the Committee's order on two grounds. Its interpretation of “workplace” was correct, and in any event Sections 2(a), 9 and 11 of the Act gave the Committee jurisdiction to entertain the complaint first and decide whether the incident occurred at a workplace as part of the inquiry itself.

The Court sided with the petitioner on both points. It read Section 2(o)(v) to mean that transportation counts as a workplace only when the employer provides it. As the judgment puts it, to fall within the meaning of workplace, the transportation has to be provided by the employer (para 21). The shared autorickshaw was public transport that neither the petitioner's bank nor the complainant's employer had arranged, so the incident did not happen at a workplace, and the Committee had no jurisdiction to entertain the complaint (paras 22 and 23). And the Court rejected the bank's entertain-first argument outright: the workplace question is a threshold jurisdictional question that an Internal Committee must answer before it acquires the power to inquire at all, not a fact it may decide along the way (paras 24 to 26).

That is a coherent reading of the words on the page. It is not the only one available, and it is worth being precise about what the Bench foreclosed and what it left standing. The judgment interprets one sub-clause of Section 2(o)(v), the one that expressly mentions employer transport, on facts involving a commute. It does not hold that “any place visited by the employee arising out of or during the course of employment” is confined to employer-arranged journeys wherever that phrase applies, and the facts did not require it to. A visit to a client's office by auto, a business trip on a public flight, a training session reached by shared cab: none of these were before the Bench, and each could still satisfy the broader language on its own terms, without passing through the transportation clause at all. What the ruling forecloses is narrower: the argument that public transport used for the ordinary two-way commute becomes a “workplace” merely because the employee happened to be travelling to or from work at the time.

Why That Precision Still Leaves a Real Gap

The narrowness of the holding does not make the consequence narrow. The daily commute is the one journey almost every working woman in urban India makes twice a day, and it is also the journey for which employer-arranged transport has become the exception rather than the rule. Where large factories and government offices once ran dedicated buses, and where a handful of employers still do for night shifts, the default now is exactly what the petitioner used: a train, an app-hailed cab, a shared auto, or some combination of all three, arranged by the employee and paid for out of the employee's own pocket. Read this way, Section 2(o)(v) puts the default mode of commuting categorically outside “workplace”. An Internal Committee then has no jurisdiction over harassment that happens on it. It does not matter that the perpetrator is a colleague. It does not matter that the journey exists only because of the job. It does not matter how routinely both parties make it.

It should be said plainly that these facts are an imperfect vehicle for that problem. The two people in the autorickshaw worked for different employers and had never met; on no reading of Section 2(o)(v) was the bank's Internal Committee an obvious forum for a dispute between them. The difficulty is not with the outcome on these facts but with the reasoning used to reach it, which sweeps well past them. The Vishaka guidelines that preceded the statute, and the Act's own definition of workplace, deliberately reached past the four walls of an office, because the Supreme Court and the drafters both understood that harassment follows the conditions of employment, not a building's boundary. The Delhi High Court made the same point in Saurabh Kumar Mallick v. Comptroller and Auditor General of India, 151 (2008) DLT 261: a narrow and pedantic approach cannot be taken in defining “workplace” by confining it to the commonly understood idea of an office. Whether a place counts as a workplace depends on the facts of the case, not on a fixed physical boundary. That line of authority was not written with the commute in mind, and it would be too much to say Parliament legislated in 2013 with app-hailed cabs in view. But its logic runs against bright lines drawn on the basis of who arranged a journey rather than why the journey was made. By treating the express mention of employer transport as the ceiling of coverage rather than one illustration of a broader principle, the Bombay High Court's reading produces exactly the kind of result that reasoning was meant to guard against: a woman's protection against a colleague's conduct now depends on who booked the cab.

The Argument That Cuts the Other Way, and Why It Does Not Fully Answer the Problem

The strongest defence of the Bombay High Court's approach is practical. Jurisdiction cannot rest on facts an Internal Committee is poorly placed to establish. A shared autorickshaw hailed off the street is outside any employer's control; an employer-run shuttle is not. A bright line between the two gives Internal Committees a workable test instead of an open-ended inquiry into how connected a commute was to the job on a given day. And here, as already noted, the alleged wrongdoer was not even a colleague of the complainant. Asking an Internal Committee to inquire into an incident between two strangers thrown together by an auto driver's overbooking is, on this view, asking it to do something the statute was never built to do.

That argument has force, but it proves less than it needs to. The line the Court drew is not between employer control and no employer control. It is between employer-arranged and employee-arranged transport, which is a proxy for control rather than the thing itself. An employer that runs a shuttle bus does exercise some control over it: it selects the vendor, it can vet the driver, it can fit cameras. But that is control over the vehicle, not over what one passenger does to another inside it, which is the only question an Internal Committee is ever asked to inquire into. On that question the shuttle and the metro are not meaningfully different. In both, the employer learns what happened after the fact, from the complainant, and inquires into it the same way. Yet only the first is captured by the literal clause. And the concern about strangers on a public conveyance, real on these facts, does not answer the harder case the same logic also forecloses: two colleagues from the same office, riding the same shared auto or the same aggregator cab every day because their commutes overlap, harassed by each other on that ride. That is precisely the dispute the pre-2013 workplace jurisprudence was written to reach. On the Bombay High Court's reading, Section 2(o)(v) now excludes it as cleanly as it excludes the stranger's case.

What Follows From Here

The fix, if there is to be one, is legislative rather than judicial. A Division Bench applying the words Parliament chose is not the forum to expand a jurisdictional definition, and the Satpute Bench cannot fairly be faulted for declining to stretch a specific transportation clause beyond what its language supports. But the 2013 Act was written for a workforce that still moved largely on employer buses, factory transport and government vehicles. Section 2(o)(v) now reads like a provision built for infrastructure most urban employees no longer use. Parliament and the Ministry of Women and Child Development, which administers the Act, have a concrete and narrow question in front of them: should coverage of commute-time harassment between colleagues depend on who paid for the ride, or on whether the ride exists because the job does.

Until that question is answered, employers have less room to manoeuvre than they may think. Jurisdiction under the PoSH Act is statutory, and Section 2(o) is a definitional gate: an employer cannot confer it on its Internal Committee by writing a broader clause into its own policy, any more than parties can agree to give a tribunal a power the statute withholds. What an employer can do is keep commute-time misconduct between colleagues expressly within the disciplinary reach of its service rules and code of conduct, so that conduct falling outside the Act is not thereby placed outside all internal accountability. That is a different and weaker instrument. It carries none of the Act's procedural architecture, none of its protections for the complainant, and none of its safeguards for the accused. It is a stopgap, not an answer, and it is worth saying so rather than pretending otherwise.

Author Kshitij Saruparia is an Advocate practicing at Rajasthan High Court & Apeksha Kachhawaha is an Advocate based in Delhi. Views are personal.