AI stock ordering puts SA franchise contracts under scrutiny

South African franchise agreements need to keep pace with automated decision-making, a new commentary argues, as artificial intelligence tools influence stock orders, staffing and pricing. Writing on 6 November, Johan Steyn raised a practical question for the franchise community: who carries the loss when a system mandated by head office gets a decision wrong?

Automation raises a question of responsibility

Steyn’s commentary, titled The Franchisor’s Algorithm Ordered the Stock. Who Pays When It Is Wrong?, examines the gap between control over a trading decision and responsibility for its financial consequences.

His argument starts with the traditional franchise relationship: the franchisor provides the brand and operating system, while the franchisee contributes capital, local knowledge and personal risk. Automated tools can alter that balance when they take on decisions previously made by the person running the outlet.

Steyn cites an Absa franchising specialist who identifies predictive demand forecasting among the AI tools owners should adopt to remain competitive. He does not reject that technology. Instead, he argues that the contractual arrangements supporting it need attention alongside its potential trading benefits.

The distinction is important: adopting a forecasting tool is not the same as agreeing who must absorb a loss if its recommendations are compulsory and prove wrong.

International developments frame the warning

The commentary draws on international examples to explain why South African franchisees should examine these arrangements before a dispute arises.

Citing law firm Greenberg Traurig, Steyn reports that franchisors are using AI for inventory and staffing forecasts, as well as monitoring individual outlets. He also points to challenges in the United States involving technology fees and mandatory vendors, where regulators and litigants have questioned alleged hidden margins and unfair practices.

Steyn separately highlights litigation brought by Chaac Pizza Northeast, an operator of about 111 Pizza Hut outlets in the United States, which is suing its franchisor for more than US$100 million. That is an overseas lawsuit, not a South African ruling or an established finding of wrongdoing.

The broader issue, in Steyn’s account, is that responsibility can become difficult to trace when several organisations participate in an automated decision. He cites Risk Management Magazine’s observation that liability may be dispersed among the model provider, the system provider and the business deploying the technology.

For a franchisee, his concern is that head office can gain greater control over ordering, pricing or scheduling while the outlet owner remains exposed to the resulting costs.

Three protections proposed for franchise agreements

Steyn recommends that franchisees negotiate three terms before signing or renewing: transparency about how an automated system reaches decisions, a right to override it using local knowledge, and loss-sharing where a mandated system demonstrably causes harm.

These are proposed contractual protections, rather than newly announced regulatory requirements.

Transparency addresses whether an owner can understand and question a decision affecting the outlet. An override provision addresses whether local judgement can take precedence over the system. Loss-sharing addresses what happens after a harmful decision, rather than leaving that question until the parties are already in dispute.

Together, the proposals reflect Steyn’s central position: the allocation of risk should keep pace with the allocation of decision-making power. His argument is not that every unsuccessful forecast creates a claim against a franchisor, but that compulsory technology deserves explicit treatment in the agreement.

Existing protections remain relevant

The commentary also points franchisees towards South Africa’s Consumer Protection Act. Steyn notes the requirement for a disclosure document at least 14 days before signing and a ten-business-day cooling-off period. He further highlights restrictions on unreasonable fees and conduct extending beyond a franchisor’s legitimate business interests.

Those provisions form part of his argument for negotiating protections; the commentary does not announce a new AI-specific rule or a South African judgment settling responsibility for automated ordering losses.

For the franchise community, the immediate message is therefore contractual rather than technological: investment in better systems should be accompanied by clarity about authority, accountability and financial exposure.

Practical takeaway: Before signing or renewing, ask which automated decisions are compulsory, whether local overrides are permitted and how proven losses will be handled. Seek clarity in the agreement rather than relying on assumptions about who is responsible.