Artificial intelligence (AI) transparency statement

In considering AI, we have adopted the OECD definition of AI, as outlined in the policy:

‘An AI system is a machine‑based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments. Different AI systems vary in their levels of autonomy and adaptiveness after deployment.’

eSafety uses and is considering further use of AI to support efficient, lawful and evidence-informed regulation. eSafety’s use of AI is intended to support staff judgement, not replace human responsibility for regulatory decisions.

The main reasons eSafety is exploring or using AI are to:

• improve the speed and consistency of internal analysis, research and document handling
• support data management, data quality checking and insight generation
• assist software development, debugging, testing and quality assurance for eSafety systems
• help staff summarise, search and organise large volumes of internal information
• improve workplace productivity and efficiency while maintaining privacy, security, confidentiality and integrity
• strengthen eSafety’s ability to understand, educate and protect Australians from, technological change in the online sectors it regulates
• protect the wellbeing of eSafety personnel and limit the need for unnecessary repeat exposure to harmful content.

In line with the Australian Government classification system for AI use, we use and consider AI across our active domain in the following usage patterns.

Usage patterns

Workplace productivity

We are currently using generative AI to improve workplace productivity for staff, including:

• helping to answer questions from staff regarding workplace policies, research findings and industry submissions and responses
• summarising and transcribing meetings
• summarising, editing or refining documents, emails, instant messages and other content
• assist in the analysis to obtain insights from datasets and systems, including identifying patterns, trends, themes and keywords
• image processing and generation
• to transparently support the design process and education/marketing campaigns, sometimes in collaboration with third parties
• use an AI-generated voice to support text-to-speech.

Decision-making and administrative action

eSafety uses AI to support decision-making and administrative activities, not to make final decisions without human involvement. It also includes systems that suggest keywords or themes to staff, who apply their discretion in accepting those suggestions. All usage is subject to human testing and review before release, including where systems may be accessed by the public.

Analytics for insights

eSafety sees benefits in using AI to assist with data and insights in the areas of data management and obtaining insights from data through interrogation and analysis. eSafety’s data and insights can influence our approach to regulation, policy and informing advice to government on legislation and regulation.

Image processing

Some eSafety staff are exposed to sensitive and harmful content that may impact their wellbeing. eSafety is exploring the use of AI to help pre-categorise images and videos to mitigate the impact of exposure to these content types.

eSafety manages risks associated with AI use through the following measures:

• AI tools and use cases are assessed for risks, including security, privacy and suitability, before approval. An internal AI policy sets requirements for responsible, ethical and secure use, including safeguards to protect the privacy, confidentiality and integrity of agency data and operations.
• Approved AI uses are monitored to ensure they continue to operate as intended and remain appropriate over time, including periodic review of performance and unintended impacts.
• AI is used to support, not replace, professional judgement, with decisions and advice remaining the responsibility of eSafety staff.
• AI use is overseen through established governance arrangements. Where issues or increased risks are identified, outputs are reviewed by staff and escalated, and eSafety can restrict, pause or stop the use of an AI system.
• eSafety requires staff to complete mandatory training on the responsible use of AI, with additional guidance and training for staff involved in AI enabled systems.
• The AI Steering Committee is committed to staying up to date and compliant with whole of government best practice and guidance.

Future AI use intentions

eSafety may explore additional uses of AI where it supports our regulatory, corporate or operational functions and aligns with our legal obligations, risk appetite and public trust responsibilities. eSafety will take advantage of evolutionary whole of government AI initiatives such as GovAI. Any future use of AI will be subject to appropriate governance, risk assessment and human oversight.

eSafety’s use of AI is governed through clear accountability and oversight arrangements:

We have established an overarching approach to AI, supported by an AI Steering Committee that oversees AI adoption, considers proposed use cases, and assesses associated risks and alignment with Australian Government policy and ethical principles.

We have appointed a Chief AI Officer (CAIO) to lead AI adoption across the organisation, champion strategic change and accelerate consistent and collaborative AI capability development across the APS.

eSafety has an AI Accountable Official to oversee implementation of the Australian Government Policy for the responsible use of AI in government and ensure appropriate governance, compliance and assurance arrangements are in place.

We will not use AI in ways that are inconsistent with our legal obligations, public trust responsibilities, or Australian Government policy. AI will not make final regulatory decisions or enforcement decisions without accountable human judgment.

eSafety will use AI where it helps us do our work better, safely and lawfully. We will use AI to support our regulatory, corporate and operational functions where it improves efficiency, supports better analysis, or reduces manual work. eSafety will consult with staff and their representatives when considering AI uses that may materially affect work practices, roles or responsibilities. 

eSafety does not currently deploy AI in direct interactions with the public. Where AI is used to create content for public education it will be accompanied by an appropriate transparency statement.

This AI transparency statement was last reviewed in June 2026. We will review and update it annually, as our AI use matures or when a significant change is made to the agency’s or the Government’s approach to AI.

For questions about this statement or further information on our use of AI, you can get in contact with us.

Contact us