E-commerce Enforcement Discussion Highlights Platform Accountability Gaps

MediaNama attended a closed-door discussion in India, held under Chatham House Rules, addressing the convergence of e-commerce, artificial intelligence, and illicit trade. The group comprised brand owners, law firms, customs and enforcement officials, intergovernmental agencies, foreign intellectual property offices, and researchers. Notably, no e-commerce platforms were present, leading to a critical, unopposed perspective.

The core complaint throughout the three-hour session was that platforms control the data, commercial incentives, and seller onboarding processes without external accountability. Seller verification, or Know Your Customer (KYC) checks, was identified as a foundational trust element that frequently fails. Speakers reported instances where registered seller addresses were non-existent, and broadly supported a shift from one-time checks to continuous, risk-weighted verification, especially for sellers with a history of counterfeit complaints.

Methods used by networks of intermediaries to reintroduce flagged sellers onto platforms were detailed, involving fabricated authorization letters and forged documents. The proposed solution included standardized onboarding checks across platforms and shared alerts to prevent sellers removed from one marketplace from reappearing on another. A significant economic point raised was that platforms earn commissions on sales regardless of whether goods are genuine or counterfeit. Sellers with infringement complaints can still access advertising and preferred placement, yet no platform policies were identified that tie enforcement outcomes to commercial privileges.

Concerns were also raised about the blurring lines between active and passive intermediaries, as platforms now employ recommendation engines, curate listings, and manage fulfillment. The argument was made that platforms actively promoting listings, potentially infringing ones, are not passive. The discussion also touched upon the risks posed by AI shopping agents, which are currently ill-equipped to distinguish genuine from counterfeit items and could be manipulated.

Three unresolved issues concerning AI and copyright in India were discussed: the validity of a recent ruling linking AI training to fair dealing and the absence of infringing output; the practical challenges of remedies once infringing material is embedded in an AI model, potentially favoring damages over injunctions; and the lack of licensing access for small and mid-sized rights holders. The organizers plan a second round of discussions that will include e-commerce platforms to address these issues.