Findings of the study

• Pages: 5-10

5

Findings of the study

General findings

1. Hotlines are predominantly reactive

 Hotlines remain predominantly reactive to reports of online CSAM received from

members of the public. A very small number of h otlines (four) engage in proactive

search (of varying f orms) for CSAM online. In the strictest sense – active search for

CSAM - only the UK’s Internet Watch Foundation (IWF), Canada’s Cyb ertip.ca and the

National Centre for Missing and Exploited Children (NCMEC) in the US do so. The

Lithuanian hotline Šva rus internetas undertakes limited and circumscribed proactive

search in the form of guided Uniform Resource Locator (URL) crawli ng.

 The volume of CSAM identified through public reporting represents the ‘tip of the

iceberg’ of the (growing) volume of CSAM that exists on both the surface and darkweb.

Reports of online CSAM rely on the visi bility of hotlines and for the public to overcome

their fear of reporting illegal material. Both act as obstacles to wid espread reporting

and reduce the effectiveness of tackling CSAM solely based on public reports. Statistics

show that proactive search leads to a substantially hig her number of identified CSAM.

In 20 18, INHOPE hotlines processed 337,58 8 alleged CSAM images reported by the

public. That same year, the UK hotline handled 114,593 CSAM items resulting from i ts

proactive search of CSAM, while the Canadian hotline’s proactive web crawling

operation identified 1.8 million CSAM items between 2016 and 2019 .4 While public

reporting l eads predominantly to the di scovery of previously unknown mat erial, pro-

active search predominantly supports remova l o f al ready known material, which is

reappearing on the internet.

2. Hotlines are constrained by their regulatory envi ronments and resource

capacity

 EU-based Hotlines are limited by what they are legally allowed to do. In practice, this

means th at most are limit ed to a reactive role (see conclusion above). Even among

reactive hotlines, there is variation in capacity: most (26 jurisdictions5) can view,

assess, localise and forward CSAM, while four jurisdictions do not allow hotlines to view

the content of the reports they receive.

 Only five EU Member States, as well as Canada and the US, have adopted legislative

acts providing a clearly defined mandate for their national hotlines. In the remaining

Member States, hotlines operate based on co-regulation or self-regulation frameworks

(such as cooperation agreements, Memoranda of Understandings (MoU) or signed

letters), which can create a lack of clarity on th e scope of their mandate, in p articular

regarding the sharing of data.

 Hotlines are further constrained by resources. Even those i nterested in conducting

proactive search, for example, are l imited by resource capacity (in addition to legal

barriers).

3. Technological developments outstrip hotline capacity

 While reactive reporting which l eads to the d etection and removal of hitherto unseen

CSAM is an important constituent of tackling CSAM, the current volumes of reporting

from internet companies suggest that they have an important role to play in supporting

hotlines to identify and remove (mainly re-appearing) online CSAM. Statistics on

reporting volumes reveal both the size of the problem and the capacity of in ternet

companies to act at scale to remove the content. In 2019, the CyberTipline in the US

4 Project Arachnid website available at: https://projectarachnid.ca/en/.

5 Including the UK, US and Canada, which were also analysed in this Study.