The ACCC is keeping an eye on the potential for anti-competitive conduct arising through big data e-collusion. Speaking at an event in Sydney last month, ACCC Chairman Rod Sims addressed the opportunities and challenges posed by the growing use of algorithms and big data.
While the ACCC acknowledges the economic advantages of data-driven innovation, it has raised concerns about the potential for such innovation to increase the risk of collusion and anti-competitive practices.
There are concerns over the way that machine learning algorithms are determining prices and achieving potentially collusive outcomes. An anti-competitive equilibrium may result if similar algorithms are being used by competing companies. In the context of mergers and acquisitions, Rod Sims noted concerns where the acquirer and target company are both involved in the collection of big data, or they are "vertically linked" in the big data supply chain. Such an acquisition could lead to the newly merged company preventing rival companies from accessing data that is necessary to compete in the market, or prevent new companies form entering the market.
This is further complicated by the fact that deep learning and artificial intelligence may mean that businesses would not know how or why their machines reach a certain conclusion. However the ACCC remains firm that businesses cannot use this as an excuse for engaging in anti-competitive practices, noting that companies cannot avoid liability by saying 'my robot did it'. The ACCC is confident that the Competition and Consumer Act 2010 is sufficiently able to deal with such conduct should it arise in the future.
The ACCC's new Data Analytics Unit is also working to monitor and prevent incidents of e-collusion. The new unit is supporting the work and investigations of the ACCC by analysing algorithms and artificial intelligence technology used by businesses.
A copy of ACCC Chairman Rod Sims' speech is available here.