Algorithms, Competition and Consumer Harm: Call for Information - Which? response

We review the CMA's report (published Jan 2021) entitled: Algorithms: How they can Reduce Competition and Harm Consumers. This is the launch of the CMA's programme of work on analysing algorithms so they can identify and better address harms. 

The CMA are very worried about:

• personalisation - hard to detect 
• manipulation of consumer choices via algorithms
• disruption to competition
• auditing of black box and machine learning programmes

Algorithmic decision making is identified as being beneficial to consumers: 

• saves us time
• gives individual recommendations
• ability to filter out a lot of the noise of the internet to help consumers focus on what's important to them

Also identified as being beneficial to business:

• helps them design better "choice architecture" by learning what people focus on and the impact on how presentation can influence decision making (though this is nudge theory and potentially disruptive to consumer autonomy)
• the mitigation of online harms
• opportunity to prevent discrimination against particular groups
• ability to detect collusion between firms to ensure competitive prices.
  

Our comments follow, below.