Paper short abstract:

Despite risks, big-tech narratives propagate that the opacity of black box systems make it impossible to regulate generative AI. This paper examines the narratives that have shaped the regulatory past of high frequency trading algorithms to stipulate new regulatory imaginaries for generative AI.

Paper long abstract:

Generative AI is transforming the digital media landscape (e.g. Jungherr 2023; Lee HK 2022). However, the future of generative AI is far from certain and depends on our capacity to imagine AI as a regulatory object and anticipate its long-term impacts and risks. While we have seen attempts at regulating AI with various intentions across the US, the EU and China, the opacity and complexity of AI is understood as a barrier to understand and regulate these systems (Ferrari et al. 2023).

While opacity may create an impasse to imagine AI as regulatory object, the regulation of high frequency trading (HRT) algorithms illustrates that it is possible for regulators to abandon the idea that an algorithm needs to be ‘opened’ in order to be regulated (Seyfert 2021).

This paper examines the narratives that have shaped the regulatory past of HRT algorithms to stipulate new regulatory imaginaries for generative AI. It does so by drawing parallels between the anticipated risks of HRT algorithms prior to the global financial crisis and present controversies surrounding generative AI. These parallels focus in particularly on the narratives of opaque systems and technological developments outpacing regulatory efforts. While doing this shows that these narratives work to supress regulatory possibilities in mitigating risks, regulatory efforts after the financial crisis reveal that these can only be suppressed for as long as risk remains uncertain. For the anticipatory governance of generative AI, it is therefore crucial to create participatory practices that illustrate anticipated risks for reflective decision making.