Ofcom consultation on protecting people from illegal harms online - Which? response

Which? welcomes this opportunity to provide our views on the proposed requirements to be placed on online services to protect UK consumers from illegal harms. Our response focuses on fraud detection measures proposed in the draft Illegal content Codes of Practice and Guidance. 

While we recognise that Ofcom’s proposals are a step in the right direction, in protecting consumers online, we do not anticipate the measures proposed for fraud detection will result in systemic change. We note that Ofcom’s fraud detection proposals have been assigned to large online and multi risk services, yet these services already have systems and practices that go above and beyond Ofcom’s proposed measures. 

In order to leverage opportunities presented in the draft Codes of Practice and Guidance to provide consumers with greater protections against fraudulent online content, we propose that Ofcom consider the following measures: 

• Calculation of Online Service Size - we recommend Ofcom consider applying the following options with the intention to capture online services such as dating sites where fraud is prevalent. 
  • 1a: Extend fraud detection duties to services with over 700,000 users.
  • 1b: Broaden the scope of risk profiles to include social context of online service.
  • 1c: Calculate large online services through its group capacity, revenue and reach.

• Use of URLs to detect fraudulent content.
  • Services with over 700,000 users should be required to use URL detection to detect possible fraud.
  • Add using unknown contact points as a risk factor in the Illegal Content Judgement Guidance.

• Future proofing fraud detection.
  • Ofcom should use its information gathering powers to generate an evidence base on the use of new technologies (including artificial intelligence) to tackle fraud.
  • Ofcom should create a standard in their codes of practice that online services use specific types of data to train their systems to prevent fraud.
  • Add Specified Anti Fraud Organisations (SAFOs) to the list of trusted flaggers.

• Future considerations.
  • Apply safety by design principles to all verification schemes.
  • Consider the risk of fraudsters exploiting boosted content.