SHARE
Basic online safety expectations
| Attachment | Size |
|---|---|
 Basic online safety expectations: key findings
|
343.18 KB |
|
Basic online safety expectations: full transparency report
|
1.9 MB |
In March 2024, the eSafety Commissioner gave a legally enforceable notice to a selection of online service providers: Google, Meta, WhatsApp, Reddit, Telegram and X; requiring them to report on measures taken to protect Australians from the risk that terrorist and violent extremist material and activity (TVE) posed to their safety and security.
This key findings report and accompanying transparency report summarise the responses Google, Meta, WhatsApp and Reddit gave to the non-periodic reporting notices, and the information Telegram provided regarding the steps being taken to detect and address TVE on their services. The reports also summarise information received from Reddit and Telegram regarding the steps being taken to detect and address child sexual exploitation and abuse (CSEA) on their services.
This document highlights some of the key findings from responses to the notice by Google, Meta, Reddit and WhatsApp. It also includes some information provided by Telegram after the deadline. The full transparency report contains additional information and context for these key findings.
Key findings
- Google reported it had some measures in place to stress-test and calibrate its generative AI service, Gemini, and to proactively prevent it from being used for harmful material and activity such as TVE and child sexual exploitation and abuse.
- Google, Meta and Reddit all reported removing individual items of TVE from their services to prevent the content from being recommended to users.
- Telegram reported that it had only one reporting option for end-users to make complaints about harmful or abusive messages in Chats and Secret Chats.
- Trust and safety staff at Reddit, Meta’s Facebook and Telegram were not automatically informed when volunteer moderators removed an account for a TVE violation.
- Although tools are being used to detect TVE – including images, videos and written material – they are not always applied consistently or comprehensively.
- Google, Meta, Reddit, Telegram and WhatsApp all reported having measures in place to address recidivism on their services.
