04 Jul 2022
Posted in Thematic Research
Big tech gatekeeping algorithms unsustainable and unethical, says GlobalData
Following a Japanese court ruling that restaurant review platform Kakaku.com unfairly altered AI-driven restaurant scores and should disclose part of its algorithms;
Emma Taylor, Thematic Analyst at GlobalData, offers her view:
“The ruling highlights the tension between big tech companies’ entitled belief that they have a right to keep their algorithms as trade secrets and regulatory bodies demanding these algorithms should be laid bare. Platforms will now be concerned that they will be forced to reveal their algorithms’ workings or risk an expensive and gruelling court case.
“Big tech has long been accused of algorithmic manipulation and secrecy. The ruling is part of a real and momentous movement demanding greater transparency. Now more than ever, it is in big companies’ interests to comply with an algorithmic transparency standard to avoid potential legal action and difficulties like strike action, which Uber has recently experienced. A desirable consequence of this case would be for it to spark a domino effect, forcing Big Tech to implement adequate algorithmic transparency regulation.”
Sarah Coop, Thematic Analyst at GlobalData, offers her view:
“Algorithms have a history of making biased decisions, and it is only a matter of time until algorithmic transparency becomes essential. Gatekeeping algorithms is unsustainable and unethical when they are used to make decisions with significant consequences.
“Companies are increasingly using neural networks with multiple layers of algorithms. Neural networks improve the accuracy of AI but also increase its opacity. Big tech companies must consider individual personal and business outcomes rather than plugging more data into large multi-layer models, which are unexplainable. Transparency and explainability must be intrinsic in AI development.”