Money. Identity. Technology.
Follow the money. Find the decision.
Regulator endorsement of particular technical approaches
The Australian eSafety Commissioner’s March 2026 Safety by Design toolkit gives industry guidance on preventing technology-facilitated child sexual exploitation and abuse. Its recommendations cover investigative AI and encrypted services, drawing on several different research sources.
The toolkit describes itself as an output of a December 2024 eSafety–US DHS workshop. Its encrypted-services case study advises considering client-side scanning, homomorphic encryption, risk-based deployment and conditional limits on end-to-end encryption. A separate footnote cites the SaferAI paper for investigative triage. The DFC report supports the discussion of platform hopping and is also linked beside the homomorphic-encryption item. Those are distinct citation relationships: neither establishes the efficacy of every technique nor identifies who drafted or approved the full recommendation list.
eSafety CSEA Safety by Design toolkit (March2026) → eSafety encrypted-service recommendations
Toolkit version displayed4March2026
Printed24/PDF13 case study lists client-side scanning, homomorphic encryption, risk-based E2EE deployment and conditional limits on E2EE. This is an affirmative guidance recommendation. Source efficacy/privacy assurances remain attributed; the SaferAI citation elsewhere does not establish origin of these recommendations.
SaferAI November 2024 discussion paper → eSafety CSEA Safety by Design toolkit (March2026)
March2026 toolkit
Printed28/PDF15 footnote34 supports classifiers categorising reports, prioritising cases and identifying behavioural patterns. Resource listing alone is not the evidence; actual footnote supports scoped use. Other age/identity/encryption sections have separate ancestry.
Guiding principles for technology-facilitated CSEA (December 2024) → eSafety CSEA Safety by Design toolkit (March2026)
Toolkit labelled v 4 March 2026; exact publication day not established; footnote access30Jan2026
Footnote29 follows the second item only; report/metadata do not identify Security and Safety vol4 as publication. This is citation placement, not verified technical support or authorship of the list.
Australian eSafety Commissioner → eSafety CSEA Safety by Design toolkit (March2026)
Version label4March2026; citation access26November2025
Toolkit describes itself as output of a December2024 eSafety/DHS workshop; paper citation accessed November2025. Do not infer the citation was present at that earlier workshop.
Guiding principles for technology-facilitated CSEA (December 2024) → eSafety CSEA Safety by Design toolkit (March2026)
Toolkit labelled v 4 March 2026; exact publication day not established; footnote26 access10Jan2025
Report p34 matches toolkit claim about movement to encrypted messaging. This separately supported contextual citation does not establish the homomorphic recommendation.
Encryption-guidance origins: funding, review access and the missing wording record ↗
SaferAI policy output: what was proposed, and where it was used ↗
Australian eSafety Commissioner → publishes guidance in → eSafety CSEA Safety by Design toolkit (March2026)
Version label4March2026; citation access26November2025
Toolkit describes itself as output of a December2024 eSafety/DHS workshop; paper citation accessed November2025. Do not infer the citation was present at that earlier workshop.
SaferAI November 2024 discussion paper → is cited to support AI investigative triage in → eSafety CSEA Safety by Design toolkit (March2026)
March2026 toolkit
Printed28/PDF15 footnote34 supports classifiers categorising reports, prioritising cases and identifying behavioural patterns. Resource listing alone is not the evidence; actual footnote supports scoped use. Other age/identity/encryption sections have separate ancestry.
eSafety CSEA Safety by Design toolkit (March2026) → advises industry to consider → eSafety encrypted-service recommendations
Toolkit version displayed4March2026
Printed24/PDF13 case study lists client-side scanning, homomorphic encryption, risk-based E2EE deployment and conditional limits on E2EE. This is an affirmative guidance recommendation. Source efficacy/privacy assurances remain attributed; the SaferAI citation elsewhere does not establish origin of these recommendations.
Guiding principles for technology-facilitated CSEA (December 2024) → is linked by footnote29 beside homomorphic-encryption guidance in → eSafety CSEA Safety by Design toolkit (March2026)
Toolkit labelled v 4 March 2026; exact publication day not established; footnote access30Jan2026
Footnote29 follows the second item only; report/metadata do not identify Security and Safety vol4 as publication. This is citation placement, not verified technical support or authorship of the list.
Guiding principles for technology-facilitated CSEA (December 2024) → is cited at footnote26 for platform-hopping context in → eSafety CSEA Safety by Design toolkit (March2026)
Toolkit labelled v 4 March 2026; exact publication day not established; footnote26 access10Jan2025
Report p34 matches toolkit claim about movement to encrypted messaging. This separately supported contextual citation does not establish the homomorphic recommendation.