A binding treaty Islamabad didn't want
On June 12, 2026, the International Labour Conference closed its 114th session in Geneva by adopting Convention No. 193, the Decent Work in the Platform Economy Convention — the first binding global labor treaty written specifically for gig and platform workers (ILO record of decision). The vote was 406 in favor, 8 against, with 36 abstentions. Pakistan was not among the eight no votes — the United States and New Zealand cast those — but it had spent the negotiating session arguing for the opposite outcome: a non-binding recommendation instead of a treaty with teeth (Geneva Solutions).
Pakistan's position aligned with the International Organisation of Employers, which called the draft convention "overly lengthy and prescriptive" and argued a stand-alone recommendation was "the most feasible path to consensus in 2026," particularly on the provisions covering algorithmic management and compensation standards (Geneva Solutions). That's not a frivolous objection. Recommendations let governments calibrate rules to local labor-market conditions without the compliance and reporting obligations a ratified convention carries, and a country with Pakistan's informal-sector footprint has real reasons to want flexibility rather than a fixed international floor.
What the convention actually requires
Convention 193 obligates ratifying states to build worker classification frameworks based on how work is actually performed and paid, guarantee timely pay and minimum wages for employees, extend social security access, and require platforms to disclose when automated systems are managing workers and to explain consequential decisions — suspensions, deactivations, non-payment (ILO, 114th ILC closing). It also folds in the ILO's core protections: freedom of association, elimination of forced and child labor, safe working conditions, and specific safeguards for migrant platform workers.
Ratification is voluntary — Pakistan is not compelled to adopt Convention 193 just because the Conference passed it — but the treaty now exists as the reference standard international labor bodies, trade partners, and domestic unions will measure Pakistani platforms against, whether or not the state formally signs on.
The gap the treaty is naming already exists at home
What makes Pakistan's recommendation-only push hard to defend on the merits is that its own domestic evidence points the other way. The second Fairwork Pakistan ratings report found that none of the country's six major platforms — Careem, Foodpanda, Bykea, Uber, inDrive, and GharPar — scored on any of the fair-work principles it assessed. No platform showed evidence that workers earned Pakistan's statutory minimum wage of PKR 28,253 (~$99) a month, let alone a living wage; ride-hailing workers in particular reported spells of negative income after platform deductions. Safety training, protective equipment, rest areas, and functioning emergency channels were largely absent, and none of the six platforms met the report's bar for engaging with organized worker representation (Fairwork Pakistan).
Thousands of Foodpanda riders in Karachi have already struck over falling per-delivery rates — a domestic signal that the informal bargain platforms have offered workers is fraying well before any international treaty forces the issue.
Provincial lawmakers have started to notice. The Punjab Labour Code 2026, notified after gubernatorial assent in late May, formally brings food-delivery, ride-hailing, and other app-based workers under labor legislation for the first time in the province — including restrictions on the non-compete clauses that had barred riders from working multiple platforms, and a push against salary-confidentiality clauses that obscure real pay (The Nation). That's a provincial statute, not federal ratification of Convention 193, but it shows Pakistani policymakers converging on the same gaps the ILO treaty targets — classification, transparent pay, algorithmic accountability — through domestic law rather than treaty obligation.
The case for proportionate, not absent, regulation
The steelman for Pakistan's position has some grounding: a one-size-fits-all binding standard, applied uniformly from Berlin to Lahore, could push cautious platforms to simply throttle expansion in lower-margin markets rather than absorb compliance costs, at real cost to the millions of Pakistanis for whom platform work is a genuine income entry point. That risk is real and worth weighing.
But the Fairwork findings show the counterfactual to regulation isn't a thriving, lightly-governed gig market — it's a market where none of six major platforms clear even a minimal fairness bar, workers occasionally go into debt to work, and safety training is treated as optional. A recommendation with no binding force does nothing to change that baseline; platforms that fail every voluntary fair-work metric today have no incentive to adopt a voluntary international one tomorrow. Pakistan's own Punjab experiment — narrow, proportionate rules on classification and pay transparency rather than a sweeping federal overhaul — is a better model than either the binding-convention maximalism Pakistan resisted in Geneva or the deregulated status quo Fairwork just documented. The Convention 193 outcome should be read less as a defeat for Pakistan's flexibility argument and more as confirmation that the flexibility Pakistan already had wasn't being used.