Artificial Intelligence Strategy 2026-28

Commissioner’s foreword

I am pleased to launch the AFP Artificial Intelligence Strategy 2026 to 2028. This strategy sets out how the AFP will harness AI to defend and protect Australia and Australia’s future from domestic and global security threats.

AI is already shaping the operating environment of modern policing. Criminals are exploiting sophisticated tools to cause harm at speed and scale. The AFP must respond with equal sophistication, guided by our values and the law, and with human judgement at the centre of every decision. This strategy commits the AFP to a human-led, responsible and transparent approach to AI. We will use AI to enhance how we investigate crime and help protect many of our members who view online child exploitation and violent extremist material in their role to keep Australians safe. AI will be a defensive and disruptive tool.

Our focus is practical and operational. We will use well-designed and tested AI systems to process large and complex volumes of information, so our members can spend more time on the work that matters most. We will invest in skills and leadership so that every member can understand, question and use AI responsibly.

The AFP will embed clear principles, safeguards and assurance across the AI lifecycle, consistent with national expectations and the standards set for policing in Australia and New Zealand.

We will be open about how we use AI wherever possible, while protecting sensitive methods and operations. Accountability for decisions always rests with the AFP.

We will build enduring partnerships across Australian policing, the Commonwealth, national security partners, state and territory agencies, industry and academia, and with our international counterparts. Collaboration will help us innovate responsibly, adopt best practice and counter the criminal misuse of AI, including deepfakes, disinformation and cyber-enabled harms.

The expectations I set are simple. AI will support, not replace, the judgement of AFP members. We will learn from experience, improve through evidence and consultation, and continually align our capabilities with community expectations.

The legitimacy of policing in Australia relies on trust, consent and fairness. This strategy reflects our commitment to those principles. By combining the expertise of our people with responsible technology, the AFP will enhance our ability to prevent harm, bring offenders to justice and keep Australians safe.

I commend this strategy to every member of the AFP and to our partners. Together we will use AI responsibly to deliver maximum impact to the criminal environment and a safer future for all Australians.

Krissy Barrett APM
Commissioner
Australian Federal Police

Our vision

Our vision is to establish the Australian Federal Police as a collaborative leader that brings law enforcement and the broader AI community together in the applied, responsible use of AI.

Executive summary

The AFP is committed to becoming a leader in the responsible application of AI to enhance the protection of Australians and Australia’s interests. This strategy outlines the AFP’s approach to adopting AI in a way that strengthens public safety, responds to the criminal misuse of AI, and upholds the trust of the community we serve.

AI presents both significant opportunities and complex challenges for policing. Criminals are already exploiting AI to commit crimes at speed and scale. In response, the AFP is adopting a human-led approach to AI that enhances our ability to prevent, detect and investigate crime, supports the wellbeing of our members, and maintains public confidence.

The AFP operates in an environment of exponential data growth. Criminal activity is increasingly cyber-enabled, generating vast volumes of electronic evidence across electronic devices, cloud platforms, chat logs and online services. Manual review of this information is no longer feasible, limiting timely response to the imposed threat as the volume and complexity is beyond human scale. To keep Australia safe, the AFP must adopt AI to ensure our policing capabilities remain timely, effective and sustainable to deal with the volume and velocity of data.

People

The AFP will focus on safeguarding the wellbeing of members by reducing the risk of continual exposure to abhorrent material. The AFP will future-proof its workforce by building AI literacy and strategic capability to ensure that operational leaders, policy officers and frontline members are equipped with the skills to understand, question and apply AI responsibly. AI will be used to support, not replace, human decision-making, so that human judgement remains central to all AI-assisted processes.

Process

AI will be governed through existing oversight mechanisms, strengthened by responsible AI governance and aligned with the Australia New Zealand Policing Advisory Agency ANZPAA AI Principles and Frameworks. The AFP will embed AI assessments into its technology governance lifecycle to ensure transparency, responsibility and accountability at every stage of adoption and use. This will preserve the trust placed in the AFP by the Australian community, as well as our regional and global partners.

Technology

The AFP will prioritise the use of well-designed, rigorously tested AI systems that are explainable, reliable and meet operational outcomes. Central to this is a strong foundation in information governance and data quality, which ensures that the data powering AI is accurate, secure and trustworthy. By elevating data governance and quality as a key enabler, the AFP will build confidence in AI services, support compliance, and maintain transparency and accountability throughout the AI lifecycle. The AFP will use these systems to achieve its priorities, including protecting vulnerable communities.

Partnerships

The AFP will continue to collaborate with domestic and international partners, including state and territory policing, Commonwealth law enforcement agencies, national security agencies, academia and industry. These formal and informal partnerships will support innovation, knowledge exchange and alignment with global best practice. The AFP will work closely with our regional partners to ensure regional security and will share experiences with forward-leaning global partners.

In summary, the AFP’s AI Strategy is designed to support the operational realities of modern policing. It ensures that AI adoption is aligned with the needs of frontline members who face increasingly complex and data-intensive challenges. By embedding responsible use of AI within our operational frameworks, the AFP will enhance its ability to respond to evolving threats while maintaining trust, agility and effectiveness.

The AFP’s AI Strategy is aligned with broader organisational priorities, including the overarching AFP Strategy, the AFP Technology Strategy 2025, and the Blue Paper: To 2030 and Beyond. The AFP recognises that the legitimacy of policing in Australia depends on transparency, accountability and responsible technology adoption. Through meaningful engagement and enduring partnerships with law enforcement, intelligence agencies, academia, industry and the public, the AFP will foster a shared commitment to effective and responsible use of AI. This strategy seeks to position the AFP to lead collaboratively in the application of AI, ensuring our capabilities remain sustainable, trusted and aligned with the AFP’s mission to protect Australians and Australia’s interests.

Strategy at a glance

The strategy brings together four enabling areas:

• People
• Process
• Technology
• Partnerships.

These areas support five strategic goals:

• Enhance productivity
• Strengthen data-driven decision-making
• Safeguard member wellbeing
• Empower creativity and workforce engagement
• Counter the criminal threat from AI.

Responsible use of AI

The AFP adopts a responsible AI approach that reflects our commitment to lawful, transparent, fair and accountable use of artificial intelligence technologies. From a policing perspective, responsible AI involves balancing legal, privacy, operational and ethical considerations to ensure AI is used in ways that uphold public safety, community trust and democratic values.

This approach is informed by a range of external standards and policies that guide responsible AI adoption across law enforcement and government. Key ones include, but are not limited to:

• The ANZPAA Responsible and Ethical AI Framework, which outlines principles such as transparency, explainability, fairness and human oversight tailored to policing contexts
• Australia’s AI Ethics Principles, which promote human-centred values, reliability, privacy and accountability in AI design and use
• The National Framework for the Assurance of AI in Government, which supports a nationally consistent approach to the assurance of AI systems
• The Policy for the Responsible Use of AI in Government, which sets expectations for lawful, safe and transparent AI use across Commonwealth agencies
• The Voluntary AI Safety Standard, which provides technical guidance for the safe development and deployment of AI systems.

The AFP acknowledges that differences may exist in the use and assurance of AI across Australian state and territory policing jurisdictions. Jurisdictional variations in AI governance and implementation will be respected and considered in the context of AI development and deployment.

Our ethical considerations are embedded within our broader commitment to responsible governance, legal compliance and community trust. This ensures that our use of AI remains aligned with public expectations and the values of Australian policing.

Strategic goals

To support the AFP’s mission of defending and protecting Australia and Australia’s future, the AI Strategy focuses on investing in science- and technology-led capability, future-proofing our workforce, building domestic, regional and global partnerships, and countering the criminal misuse of AI, particularly against vulnerable communities. We will use AI to ensure our policing capabilities remain effective, trusted and future-ready.

Enhance productivity

AI will be applied to streamline workflows, automate routine tasks and reduce cognitive burden on AFP members. This includes data transformation and enrichment services such as transcription, translation, summarisation, video analytics and entity recognition, allowing officers to focus on complex decision-making and frontline engagement.

Strengthen data-driven decision-making

AI will support early identification and disruption of criminal threats through AI-assisted triage and data analysis across complex datasets, including cybercrime, child exploitation, counter-terrorism, foreign interference and transnational organised crime. By enabling faster analysis of large and diverse datasets, AI will help the AFP act decisively to prevent harm.

Safeguard member wellbeing

AI will assist in reducing exposure to psychologically harmful material, such as child abuse material, by enabling automated filtering and processing. This supports the AFP’s commitment to the health and wellbeing of its workforce.

Empower creativity and workforce engagement

AI will be applied to support problem-solving, encourage innovation and enable AFP members to focus on more engaging and meaningful tasks. By reducing administrative burden and enhancing access to insights, AI will help boost morale, increase job satisfaction and foster a culture of continuous improvement.

Counter the criminal threat from AI

The AFP will develop and deploy capabilities to detect and disrupt the criminal misuse of AI technologies. This includes countering deepfakes, synthetic media and disinformation campaigns that threaten public safety, evidential integrity and community trust.

AI principles

The AFP has fully adopted the guiding principles developed by ANZPAA in the Responsible and Ethical Artificial Intelligence Framework, endorsed by Australian and New Zealand Police Commissioners. The framework provides a foundational reference for law enforcement agencies seeking to adopt AI in a way that is lawful, proportionate and aligned with community expectations.

Explainability

AI systems must be understandable and open to challenge, with their purpose and function clearly described.

Accountability

Police remain fully responsible for all AI systems they use, including those developed by external vendors.

Human oversight

AI must support, not replace, human decision-making, with appropriate oversight at every stage of its use.

Fairness

AI systems must be designed and used in ways that uphold fairness, equality and human rights, avoiding unjust harm or discrimination.

Transparency

Information about AI use should be made publicly available wherever possible, without compromising policing objectives.

Skills and knowledge

Police members must have the right training and understanding to safely and effectively use AI systems.

Proportionality and justifiability

AI should only be used when it is reasonable, necessary and lawful, with clear benefits to community safety.

Reliability

AI systems must be continuously monitored and updated to ensure they remain accurate, relevant and fit for purpose.

Privacy and security

AI design and use should comply with relevant privacy, security, legal and data obligations.

AI risks

The adoption of AI in policing presents strategic opportunities alongside complex risks. To ensure responsible and sustainable implementation, the AFP recognises the following key risk areas and will actively manage them through structured governance, oversight and continuous review.

AI system failure risk

AI systems can fail to deliver intended outcomes due to weaknesses in design, testing, deployment, user adoption, data quality or ongoing monitoring. This can include inaccurate outputs, invented or misleading content in generative tools, performance degrading over time as data changes, or unintended consequences in an operational context. The AFP will manage this through appropriate lifecycle oversight, adherence to AI safety standards, and continuous evaluation to ensure systems remain reliable, explainable and fit for purpose.

Overuse, inappropriate or reckless use

Excessive or poorly considered use of AI may undermine operational integrity, lead to over-reliance on automation or result in decisions that lack human judgement. The AFP will mitigate this through training, uplift in AI literacy and a clear Human-AI Collaboration Framework that ensures AI supports, not replaces, human decision-making, particularly in high-risk or sensitive operations.

Legal, ethical and privacy risk

AI use must comply with legal obligations and uphold ethical standards. Risks include unlawful data processing, inappropriate release of personally identifiable information, bias and lack of transparency. These will be addressed through structured governance, privacy safeguards and adherence to responsible AI principles.

Reputational and public trust risk

Misuse or misunderstanding of AI in policing can erode public confidence. The AFP will manage this risk through transparent communication, community engagement and a commitment to fairness, accountability and human rights.

Managing bias and hallucination

AI systems present unique risks related to bias and hallucination that require careful management in policing. This can result in discriminatory outputs or reinforce negative stereotypes, especially affecting minority and marginalised communities. Additionally, AI systems may generate inaccurate information, sometimes called hallucinations, which can affect operational decisions and public trust. The AFP will manage these risks through a human-led approach, robust technical standards and ongoing oversight across the AI lifecycle, ensuring that all AI-supported activities remain fair, accountable and responsible.

AI adoption pillars

These pillars form the foundation of the AFP’s approach to responsible AI adoption and integration.

1. People

Future-proof our workforce so that our people understand AI, lead with integrity and embrace innovation, while benefiting from safeguards for their wellbeing. This includes uplifting AI literacy across all levels, fostering a culture of responsible AI practices and equipping leaders to make informed decisions. The AFP will prioritise the health and safety of its workforce by reducing exposure to psychologically harmful material through the responsible use of AI. Human judgement will remain central to all AI-supported processes, with clear guidance on when and how AI should assist, augment or defer to human decision-making.

1.1 AI knowledge, culture and leadership

• Build a workforce that understands AI, embraces innovation and leads with integrity.
• Identify opportunities to use AI to support inclusion, psychological safety and cultural safety.
• Uplift AI literacy across all levels.
• Build in-house expertise in artificial intelligence, machine learning and foundational model operations technology.
• Ensure leaders are equipped to make informed decisions about AI use.
• Build a technical workforce that continues to develop knowledge and capability in the development and deployment of AI tools.

1.2 Human-led decision-making

Ensure AI supports, not replaces, human judgement. AFP members must retain control over decisions, especially where rights, freedoms or public safety are at stake. Human oversight must be embedded in all AI-supported processes, with clear escalation and accountability pathways.

1.3 Safeguarding member wellbeing

Protect the psychological wellbeing of AFP members by reducing exposure to harmful content and ensuring AI is used in ways that support health, safety and resilience across the workforce.

2. Process

Establish strong governance and operational alignment to ensure AI is adopted responsibly and effectively across the AFP. This includes embedding safety guardrails, legal compliance and oversight into every stage of the AI lifecycle. The AFP will lead by developing and deploying AI technologies through due diligence, ensuring that all systems are thoroughly assessed before implementation. Planning, risk management and continuous review will be central to maintaining public trust and operational integrity, which is crucial to defending Australia’s domestic security.

2.1 Responsible AI governance and policies

Establish and maintain structured governance frameworks to oversee the development, deployment and monitoring of AI systems. This includes AI principles review, legal compliance, privacy safeguards and clear accountability for decisions informed by AI.

2.2 Operational alignment

Ensure AI adoption is driven by operational priorities and business needs, not by technology alone. AI initiatives are developed and delivered through a collaborative, cross-functional environment that brings together operational teams, technical experts and governance leads. This approach ensures that AI systems are aligned with real-world policing outcomes, are practical to implement and deliver measurable impact on public safety and organisational effectiveness.

3. Technology

Ensure the AFP has access to scalable, secure and fit-for-purpose technology and data environments that enable responsible and effective AI adoption. This includes sustained investment in the technical workforce, computing infrastructure and analytical tooling, and a comprehensive uplift in how data is collected, stored, enriched and analysed. This is to maximise the value of AI across the full data lifecycle and support strategic, lawful and efficient policing outcomes. With criminals increasingly using technology to harm vulnerable communities, the AFP must invest in expertise to balance the scales.

3.1 Invest in technology

Strengthen the AFP’s ability to deploy and scale AI technologies by investing in computational infrastructure, data storage and flexible procurement models that support innovation and operational resilience.

3.2 Improve end-to-end data management

Contribute to the AFP’s transformation of data environments by embedding robust information governance and data quality practices. This will enable consistent, secure and scalable management of operational data across its lifecycle, from collection and storage to enrichment, analysis, disclosure and dissemination. Elevating data governance and quality as strategic foundations ensures that AI systems are built on reliable, unbiased and auditable data, supporting lawful and responsible policing outcomes.

4. Partnerships

Strengthen collaboration across law enforcement, intelligence, academia, industry and the public to support responsible AI adoption. The AFP will continue to build enduring partnerships domestically, in our region and globally, promoting innovation, capability uplift and alignment with best practice. Through transparent engagement and shared standards, the AFP will foster trust, encourage knowledge exchange and ensure AI systems are developed and deployed in ways that reflect community expectations and uphold democratic values.

4.1 Partnership and collaboration

Work with law enforcement, intelligence, academia and industry to share knowledge, align standards and co-develop AI capabilities. These partnerships will support innovation, evidence-based development, access to emerging research and co-design of AI solutions tailored to policing needs.

4.2 Public trust and transparency

Build and maintain public trust in the AFP’s use of AI through transparent engagement, responsible guardrails and a commitment to fairness, accountability and human rights. The AFP will communicate openly about how AI is used in policing, while protecting sensitive methodology and operations, to ensure its systems reflect community expectations and uphold democratic values.

AI governance and policy

AI governance and policy consists of three connected elements:

• AI policy: responsible use.
• AI governance: responsible adoption.
• AI safety standards: technical standards.

Measuring success

To ensure the AI Strategy delivers meaningful outcomes and remains responsive to operational needs, the AFP will establish mechanisms to measure progress and evaluate impact across all strategic pillars. This includes tracking how AI adoption contributes to public safety, member wellbeing, operational effectiveness and community trust.