The Australian Competition and Consumer Commission (ACCC) has provided an overview of its approach to potential future cases where machine learning algorithms are deployed as a tool to facilitate conduct that may contravene competition law.
While the ACCC sees many economic advantages in "data-driven innovation" -- such as consumers being able to compare products online -- it also has a number of concerns, such as the possibility of such innovation increasing the risk of engaging in and sustaining collusion, and decreasing competition in the market without necessarily violating any competition laws.
"Cases brought to date globally by competition authorities relating to the use or misuse of online databases to determine prices reflect circumstances where 'something more' occurred," ACCC chair Rod Sims said at a conference in Sydney on Thursday.
In the United States Airline Tariff case, a database that was accessible to travel agents was being used by airlines to negotiate supra-competitive airfares and ensure proposed price hikes were maintained, Sims said.
"It is said that a profit-maximising algorithm will work out the oligopolistic pricing game and, being logical and less prone to flights of fancy, stick to it," he said.
The development of deep learning and artificial intelligence (AI) could also mean that companies are unaware of how or why a machine comes to a particular conclusion, he noted.
"To this end, it is argued that if similar algorithms are deployed by competing companies, an anti-competitive equilibrium may be achieved without contravening competition laws," Sims said.
On the other hand, it has been argued that increased price discrimination will make collusion more difficult because predictions require a combination of multiple elements, including limited brand differentiation.
"They say there is no evidence at this stage that the collusive result would be any worse than the current status quo in oligopolistic markets," Sims said.
In response to those who think "algorithmic collusion" is a get-out-of-jail-free card, Sims said companies cannot avoid liability in Australia by saying 'my robot did it', further referring to a 2013 case that centred on the use of Google Adwords to generate misleading search results.
"In that case, the ACCC was concerned about misleading search results produced in what were then known as Google's 'sponsored links'. These arose between 2005 and 2008 when businesses added their competitors' names to Google's advertising algorithm to ensure that when a consumer searched for, for example, 'Harvey World Travel', the consumer saw a sponsored link to 'STA Travel' at the top of the search results," Sims explained.
The case went to the High Court, which ruled that Google was not liable for the conduct, but agreed that the sponsored links were misleading. The companies that used Adwords to create the misleading sponsored links were found to have violated consumer laws, Sims pointed out.
Upon the closure of the case, Google made sure the practice no longer occurred worldwide, as it could have been found to be liable if an allegation had been made around Google being 'knowingly concerned'.
Sims also posed the question of what would happen if an AI robot engaged in collusion with another robot, either through the "predictable agent" or "autonomous machine" scenarios conceived by Harvard University law professors Ariel Ezrachi and Maurice E Stucke.
While he didn't have an answer, Sims said he is "confident" that the addition of the new concerted practices provision under Section 45 of the Competition and Consumer Act 2010, as well as the new misuse of market power provisions, can be used in response to either scenario if there is compelling evidence of anti-competitive conduct.
To properly assess competition issues, Sims said the ACCC is building expertise to analyse algorithms, having established a Data Analytics Unit to support investigation teams.
"At this stage, the ACCC has not seen any anti-competitive algorithms which require an enforcement response beyond what is now available to the ACCC under Australian law," he said.
"It is very likely, however, that big data and algorithms will expand the activities that give rise to concerted practices."
Sims went on to present a hypothetical scenario wherein a machine learning algorithm deployed by a company with substantial market power helps it determine profit-maximising downstream prices and engage in margin squeeze. In such a case, the new misuse of market power provision is "fit-for-purpose to prohibit this conduct", he said.
Meanwhile, the concerted practices provision should encourage competing companies to consider whether cooperation substantially reduces competition.
"If robots are colluding, this provision will help us stop this conduct," Sims said.
In the context of mergers and acquisitions, Sims said if the acquirer and acquired company are both involved in the collection and sale of big data, or they are "vertically linked" in the big data supply chain, it could lead to the newly merged company foreclosing access to data that is needed for rival companies to compete, or for new companies to enter the market. But he said this scenario would need to be assessed in context.
"As another example, it could be argued that the acquisitions of a maverick firm may be more harmful in a market where big data would otherwise facilitate coordination," he added.
Regardless of the nature of the anti-competitive conduct, Sims said the ACCC now has a "legal hook" to take appropriate enforcement action.
Charging fees for data access, introducing a strong liability regime in the event of data breaches, and establishing data privacy and security standards are among the recommendations provided by banks and fintechs ahead of the impending open banking regime.
The Business Council of Australia opposes mandatory data sharing in Australia, saying it will lead to companies piggybacking off the banks and other established financial institutions, rather than innovating.
While not legally binding, the guide is a step towards ensuring IoT device and service providers and consumers are working together to mitigate privacy and security risks associated with IoT.
How machine learning creates demand for human workers (TechRepublic)
Artificial intelligence and IoT are increasing demand for human workers, says Calabrio CEO Tom Goodmanson, and they're more stressed out than ever.
According to Deloitte research, smartphone users are regularly leveraging artificial intelligence and machine learning, which could make the technologies easier to market.