By James Konstantin Galvez ( May 6, 2026, 07:32 GMT | Insight) -- The Competition and Consumer Commission of Singapore said companies should remain responsible for foreseeable antitrust harms caused by artificial intelligence algorithms, regardless of whether they develop the software internally or use third-party systems. CCS Chief Executive and Commission Member Alvin Koh said businesses deploying AI pricing and decision-making tools must ensure compliance with competition law as regulators globally assess risks linked to algorithmic collusion.The Competition and Consumer Commission of Singapore, or CCS, said companies should remain responsible for foreseeable harms arising from artificial intelligence algorithms, regardless of whether they develop the software internally or use third-party systems, as regulators assess the competition risks linked to AI-driven pricing tools.Speaking at an international conference* on Wednesday, Koh said businesses deploying AI systems need to ensure their conduct complies with competition law even as companies increasingly automate pricing and commercial decisions using algorithms.“[But] my own view is that firms should remain responsible for foreseeable harms caused by algorithms, whether or not you are a supplier of software, whether or not you are, in fact, a user of the software,” Koh said.Koh added that companies “need to have a very careful look” at whether their AI systems comply with competition law, adding that “compliance by design when it comes to AI and antitrust issues” is “certainly something that we should be thinking about quite carefully.”Koh said businesses are increasingly using AI and algorithms to automate pricing and commercial decisions as generative AI systems lower technical barriers for companies, including micro and small businesses.“Businesses today are increasingly making use of AI and algorithms make pricing decisions,” Koh said, adding that the commercial availability of generative AI systems has “lowered the technical barriers for businesses to automate these commercial decisions, such as pricing.”He said AI systems can improve efficiency through real-time analysis of large datasets, quicker responses to supply-and-demand fluctuations and enhanced allocative and productive efficiencies.“You get real-time analysis of massive datasets for immediate market insight,” Koh said. “You get instantaneous response to supply and demand fluctuations.”At the same time, Koh said AI systems create enforcement challenges because of their “black box” nature, making it difficult for regulators to understand how algorithms generate recommendations or whether they facilitate collusive conduct.“The advice, the recommendations and the overall intentions of such AI, they’re really opaque and uninterpretable,” Koh said.He added that algorithmic collusion may be harder to detect than traditional cartels because authorities may no longer find conventional evidence such as direct communications between competitors.“You’re not going to have a smoking gun between human actors,” Koh said.The CCS earler launched an AI market study examining potential competition and consumer-protection issues arising across Singapore’s AI value chain. Koh said the study aims to provide businesses with guidance on potential competition risks linked to AI deployment and ways to mitigate those risks proactively.Koh said preliminary findings from the study suggest there is currently limited evidence of algorithmic collusion in Singapore.“We don’t think that there is a great deal of evidence that algorithmic collusion is a significant risk in Singapore,” Koh said.Businesses in Singapore do not yet rely heavily on third-party AI pricing applications and authorities have not identified pricing strategies suggesting collusive behavior, he said.Koh identified three broad categories of algorithmic collusion concerns: explicit collusive agreements supported by AI systems; “hub-and-spoke” arrangements involving shared pricing software; and autonomous tacit collusion, where algorithms independently choose not to compete on prices as a profit-maximizing strategy.Discussing international developments, Koh cited the US case involving rental pricing software provider RealPage, where authorities alleged the company’s software enabled landlords to align rental prices using competitively sensitive data. He also referenced UK cases involving online sellers and energy suppliers that used pricing algorithms and software tools to support anti-competitive conduct.To address emerging risks, the CCS recently introduced an open-source AI markets toolkit intended to help AI developers and deployers assess whether their systems and business practices comply with Singapore’s competition and consumer protection laws.“This is open-source software,” Koh said, adding that businesses can use the toolkit to self-assess AI models and business practices for potential competition and consumer protection issues.Koh said businesses using the toolkit could receive reduced financial penalties if they are later found to have infringed the law.*ICN Annual Conference 2026, International Competition Network, Manila, May 5-8, 2026 Please email editors@mlex.com to contact the editorial staff regarding this story, or to submit the names of lawyers and advisers....