Protecting Consumers in the Age of Algorithms
Many of us have tried to book a hotel room online only to be met with a flashing red alert warning, “only one room is left”, often amplifying non-existent scarcity and urgency. Chatbots in online food delivery platforms suggest “popular” options and combinations, nudging you to spend more. Dynamic search-engine ranking determines which results appear most prominently, often favoring items boosted through paid promotions. On social media platforms, targeted adverts are our second shadow.
These are inexhaustive examples of how digital platforms utilize Artificial Intelligence (AI) to influence consumers’ purchasing decisions.
But not everything AI is gloomy. The technology is positively used in credit scoring and in agriculture, where it supports better resource management and productivity. In healthcare, AI-supported diagnosis leads to better outcomes. Students are enjoying personalized e-learning while teachers have automated previously laborious administrative tasks.
UNCTAD projects the global AI market to surge from $189 billion in 2023 to $4.8 trillion by 2033, driven by rapid adoption and technological advances that is making AI more accessible for businesses. Globally, 48% of businesses currently utilize AI, and over 83% consider it a top strategic priority.
The Kenya Artificial Intelligence Strategy launched in March 2025 is laudable. It seeks to position Kenya as a leader in AI technology adoption, anchored on three pillars; building accessible and modern infrastructure; data and governance; and promoting research, innovation and commercialization of AI tools.
It is therefore timely that this year’s World Competition Day, commemorated annually on December 5, is themed: “Artificial Intelligence, Consumers and Competition Policy.” It acknowledges AI tools present immense benefits, but that its adoption is not without risks, including competition and consumer protection concerns.
From a competition law perspective, AI can enable businesses engaging in anticompetitive conduct to further conceal their illegal conduct. Unlike overt evidence of cartel conduct decoded through WhatsApp messages, email trails or paperwork, algorithmic tools facilitate opaque forms of collusion. This includes competitors deploying similar algorithms to track and automatically respond to each other’s prices, leading to stable but higher costs for consumers.
AI reinforces tacit coordination where firms independently monitor their rivals’ pricing strategies and deploy technology to automatically adjust theirs, sustaining cartel outcomes even without direct communication. Additionally, AI can reinforce existing market power by leveraging dominant firms' vast proprietary datasets that impede smaller competitors from accessing the market.
From a consumer protection perspective, abuse of AI tech results in negative outcomes. AI algorithms have reduced pricing clarity, as they analyze a consumer's data in the background, leading to consumers paying different prices for the same product or service, therefore buttressing inequality. As already exemplified with online hotel booking practices, AI facilitates dark pattern marketing to aggressively manipulate consumer behaviour. Additionally, AI may lead to privacy breaches, fraud and impersonation.
These concerns require a multi-pronged intervention.
Firstly, there is need to adopt and promote best practices to address algorithmic transparency, fairness, and accountability, including developing ethical guidelines and harmonizing regulatory frameworks across Government. Improved coordination between implementing agencies with cross-cutting mandates is therefore critical.
It is for this reason that the Competition Authority of Kenya has proposed amendments to the Competition Act to better attend to anti-competitive practices in digital markets. The proposed changes to the law are aligned with the aforementioned AI Strategy. The Office of the Data Protection Commissioner plays an equally critical role in data sovereignty and ensuring the datasets used to train and operate AI models is collected, processed and stored lawfully.
Secondly, regulators must invest in tools and systems that are as intelligent as those they are trying to oversight. For instance, the Netherlands Competition Agency has developed a web scraper with automatic data capture and analysis features, easing its evidence gathering capabilities. Another tool scans telemarketing phone calls to detect wrongdoing. Moreover, market studies to understand the inner workings of these digital markets should be prioritized. Regulatory sandboxes that facilitate controlled experimentation, stakeholder engagement, and early flagging of AI risks should also be explored.
Thirdly, Kenyan consumers should be empowered to identify and avoid AI-driven manipulative tactics, and assert their rights when businesses overstep or engage in unfair practices. This necessity is proven by the fact that, even with robust consumer and data protection laws, Kenyans’ personal data is still illegally accessed and processed across various platforms.
Lastly, AI and digital markets transcend borders. Competition and consumer protection regulators must foster cooperation by engaging in partnerships to share knowledge, tools and best practices, to adopt coherent AI regulations and unified consumer protection strategies.
For feedback, email: info@cak.go.ke