Preventing the spread of untruths via AI deepfakes

Ava Rawlings, Curtin University, AI For Good Summit 2026

Executive Summary

The spread of misinformation and disinformation through AI manipulated images and videos has the potential to cause severe harm to the Australian community. Online content can be manipulated to distort the truth and portray non-existent realities. When consumers of online media are misled by these distorted truths, misinformation and disinformation can spread. The likeness of public figures, politicians and world leaders can be manipulated to say or act in an unseemly manner, interfering with elections and undermining democracy. The publication of deepfake images and videos which convey untrue information has implications for the electoral process, reputations, education and trust in the media.

To help address this, there must be government action. The Australian Government should amend the Online Safety Act 2021 to give the eSafety Commissioner powers to compel internet service providers to block material that threatens public safety. This would empower the eSafety Commissioner to remove harmful deepfake content and enforce penalties for individuals, companies and internet services that publicize and spread misinformation and disinformation. When misinformation and disinformation threatening public safety is spread online that qualifies as sufficiently harmful to public safety, the eSafety Commissioner may compel the internet service provider to remove the content. 

Problem Identification

Artificial Intelligence (AI) technology has developed rapidly, making it progressively more challenging to identify real content from AI generated deepfakes. A deepfake is an online video, photo or sound of a person that falsely portrays them saying or doing something they didn’t say or do, creating a realistic but untrue depiction through editing (eSafety Commissioner, 2022). The ability for AI to be used to alter an image or video can result in AI generated misinformation and disinformation spreading (Australian Human Rights Commission, 2024).

In March 2022 a deepfake video was spread on social media and published on a Ukrainian news website before being removed, depicting Volodymyr Zelenskyy, the President of Ukraine, calling for Ukrainian troops to surrender. Meta and Youtube subsequently removed the deepfake video due to violations of misleading policies. The Australian Human Rights Commission recognised this incident as an example of how synthetic content can undermine democratic processes (Australian Human Rights Commission, 2024). There is a risk of AI deepfakes being used to spread misinformation and disinformation on politicians and government figures in Australia to manipulate election outcomes, which would cause harmful impacts to Australian democracy. This is a global problem, as the spread of misinformation and disinformation can occur internationally with AI available worldwide.

Context

Background

Current regulations do not effectively mitigate against the risk of AI being used to generate and manipulate images and videos to spread misinformation and disinformation in the form of deepfakes. The risk of synthetic media being abused is a grievous cyber threat to organisations, and society’s trust in online content is declining fast (Australian Signals Directorate, 2025). 75% of UN Peacekeepers have reported misinformation and disinformation impacting their safety and security, reflecting the great scale of concerns (United Nations University, 2023).

Misinformation can be spread without malice, while disinformation refers to misinformation intending to deceive others (Australian Government Office of Impact Analysis, 2024). Australians may suffer negative implications of the spread of misinformation and disinformation including damage to their health, vilification of societal groups, financial harm and a lack of integrity of the electoral processes (Australian Government Office of Impact Analysis, 2024).

AI manipulation can also disrupt journalism and present additional challenges  to maintaining truth and trust of the general public (AI Equity Lab, 2024). Disinformation has been deployed to manipulate the opinion of the public, discredit people and spread confusion at critical times such as elections and crises, threatening digital trust (Sudhakar Tiwari, 2025). As such, the inability for  consumers and journalists to distinguish real and fake content only exacerbates an already eroded sense of trust (AI Equity Lab, 2024). 

Current Policy Landscape

The Australian Government’s National AI Plan, released December 2025, is aimed at growing the Australian AI industry, using largely technology-neutral legal frameworks to manage this technology in the short term (Department of Industry, Science and Resources, 2025). The Australian Government’s National AI Plan has three goals; capture opportunities, spread the benefits and keep Australian’s safe (Department of Industry, Science and Resources, 2025). The implementation includes developing the AI Safety Institute, reducing online harms and providing guidance for AI adoption. There aren't any AI-technology specific statutes or regulations in Australia as existing frameworks and legislation are relied on to mitigate AI harms (White & Case, 2025). 

From 2026, an AI Safety Institute will also monitor the technology while the government continues to implement the National AI Plan. With this purpose of expanding Australia’s AI industry, strict AI regulations haven’t been proposed due to concerns this will limit the development of this technology in Australia (Craig Subocz, 2025). The eSafety Commissioner currently has powers under the Online Safety Act 2021 (Cth) to issue blocking notices to internet providers to enforce the removal of abhorrent violent content, severe cyber abuse and child sexual abuse material (Online Safety Act 2021 (Cth)). However, this provision doesn’t have scope to enforce the removal of AI generated and manipulated deepfake content which can spread harmful misinformation and disinformation.

International Policy Landscape

The European Union AI Act 2024 is the first legal framework on AI regulation, creating an example for the rest of the world to follow. This Act outlines risk-based rules for AI developers and deployers, to create cohesion across the European Union (European Union, 2025). Additionally, the act enforces obligations on deployers and users of AI systems, such as mandating transparency where AI has been used in generating content (European Union, 2025). Companies who don’t comply with the Act may face fines ranging from 7.5 to 35 million euros, or 1-7% of the company's annual turnover (Konrad Meier, Roger Spichiger, 2024). 

The regulations involve prohibiting AI systems that pose a threat to safety, rights of people and their livelihoods, such as banning publishing material that is harmful AI-based manipulation and deception and harmful AI-based exploitation of vulnerabilities (European Union, 2025). Included in the act are prohibitions against placing on the market, putting into service of use of an AI system that exploits vulnerabilities of a person due to age, disability, of social or economic situations, with the objective of materially distorting the behaviour in a manner likely to cause harm (European Union, 2025). While the policies of the AI act are viewed as being an important step in advancing AI regulation through prohibited practices, there are concerns regarding the effectiveness of transparency responsibilities as the onus is placed on the providers and deployers of AI systems to comply (Mateusz Kabuz , 2024). However, banning harmful deepfakes reduces the potential harms from misinformation and disinformation. 

Policy Options

To reduce the spread of misinformation and disinformation by AI deepfakes in Australia, the Australian Government needs to reduce the prevalence of deepfakes and mitigate their harmful impacts. New regulations are needed to reduce the prevalence of AI generated images and videos replicating untrue or inaccurate news content and increase the identification of AI generated images and videos. 

Option 1: Amend the Copyright Act 1968 (Cth) to include the protection of an individual’s likeness

A policy solution to address the spread of AI generated misinformation and disinformation through deepfakes is to grant individuals the right to the copyright of their own likeness. This has been done in Denmark and would make it illegal for photographs and videos with the likeness of people and their appearance or voice to be imitated and shared without consent (James Brooks, 2025). This will reduce the spread of misinformation and disinformation by making it a copyright infringement for AI to manufacture content of people in the media, journalists and public figures. This is proposed to be implemented at the Commonwealth level, but will contend with the issue that social media and the online web is international, thus illegal content can still be accessible globally (Law Society Journal Online, Cat Woods, 2025). 

In Australia, the Copyright Act 1968 can be amended to include the right to the copyright of one’s likeness. These amendments will need to be passed by the Australian Parliament. The main costs associated with government legislative amendments are the administrative processes, including research, legislative drafting, legal review and consultation (Australian Government Office of Parliamentary Council). This amendment can be integrated into the Australian Government’s AI Plan. However, establishing standards for what qualifies as using someone's ‘likeness’ is difficult, leading to problems when trying to regulate and measure a breach of this rule (Makayla Muscat, 2025, Ad News). 

Option 2: Amending the Online Safety Act 2021 to expand the powers of the eSafety Commissioner to order the blocking of AI manipulated deepfake material that threatens public safety.

The Online Safety Act 2021 may be amended to include a provision allowing the eSafety commissioner to request material that threatens public safety to be removed. A provision under s95(1)(b) may be added providing for the removal of material if the eSafety Commissioner is satisfied the material “is generated or manipulated by artificial intelligence and distorts the truth”. This amendment will ensure action is able to be taken quickly when content that’s manipulated by AI spreads misinformation. The purpose of this legislation is to prevent the spread of misinformation caused by the harmful AI manipulation of content (deepfakes), where harm can come to the Australian community. For example, if AI manipulated deepfake material displayed the Prime Minister issuing a media statement which was harmful and didn’t actually occur, the eSafety Commissioner would have powers to block this damaging content. 

These amendments will need to be passed by Parliament and non-compliance would result in fines. This amendment to the Online Safety Act 2021 of the expansion of the powers would align with the Australian Government’s AI Plan goal to regulate, prevent and mitigate harms through existing legislation and update policy where necessary (Australian Government Department of Industry, Science and Resources , 2026).  However, arguments against a policy like this include the hindrance to freedom of speech. The regulator must balance this risk with the benefits of the removal of content (Noelle Martin, 2025). This power may not adequately deter or prevent further online harms as it is a reactive measure (Noelle Martin, 2025). 

Option 3: Amending the Competition and Consumer Act 2010 (Cth) to include a dedicated provision mandating AI generated and manipulated content be clearly labelled and detectable.

A policy solution to address the issue of misinformation and disinformation being spread is amending consumer law to include a provision on AI generated material. These amendments would require image and video content generated or manipulated by AI to be labelled as AI generated. This will enhance content transparency, as is being implemented in the European Union where AI generated content must be labelled clearly and be detectable by machines as AI generated content (Natalia Garina, 2026). In Australia, it’s advised for people to include a label when they use AI to manufacture or change content, but this is voluntary (Department of Industry, Science and Resources, 2025).  By enforcing this as a law, recourse and punitive action would be enabled. 

These amendments could be developed and implemented by the Australian Competition and Consumer Commission and will help reduce the spread of misinformation and disinformation. This will be done by informing consumers of the content’s authenticity and preventing false information from being reproduced and portrayed as truth. This will help mitigate the risk to individuals for their likeness being used and manipulated by AI to spread false information (European Commission, 2026). The punitive element of this policy option would involve fining individuals and companies that don’t label AI generated content, similarly to the European Union AI Act, which would be upheld by the Australian Competition and Consumer Commission. If Australia continues with its existing regulatory plan and framework, amending the Competition and Consumer Act may help alleviate the spread of misinformation while aligning with the current strategy and following Europe’s lead. The enforceability of this option is in issue as this option doesn’t provide a government body with the power to remove or investigate suspicious or alleged un-labelled AI content, with the deterrent being financial penalties.

Policy Recommendation

Option 2 “amending the Online Safety Act 2021 (Cth) to expand the powers of the eSafety Commissioner to order the blocking of AI manipulated deepfake material that threatens public safety,” is recommended as the most viable way forward to reduce prevalence of harmful deepfake content. This option aligns with steps already taken by the Australian Government pursuant to the National AI Plan thus working within existing framework to minimize additional costs, time and resources. 

Implementation

The Australian Government would need to prepare amendments to the Online Safety Act 2021 s95(1)(b) to include an additional provision, giving the eSafety Commissioner the power to demand an internet service provider to remove AI generated or manipulated material provided that the material threatens public safety. The purpose of this is to expand the circumstances in which the eSafety Commissioner may use its powers pursuant to the Online Safety Act 2021 to demand the removal of harmful deepfake material when the deepfake is spreading dangerous misinformation. It is estimated this policy option would cost $8 million over a period of four years, due to similarity with the Commonwealth Government Misinformation Bill 2024, which provided an estimate of costs from 2023-2024 (Federal Budget 2023-2024). However, costs associated with the implementation of this policy may be offset by revenue received from fines. 

This proposal suggests the below insertion:

s95(1)(b)

(v) The eSafety Commissioner may demand the removal of material that it deems to threaten public safety if satisfied the material is generated or manipulated by artificial intelligence and distorts the truth. 

Key Elements

Due to the ambiguity of terms in this new provision, the following must be considered for implementation of the new provision:  

Form

The above provision would be inserted into the Online Safety Act s95(1)(b) to extend the powers of the eSafety Commissioner to facilitate the demand of internet providers to remove material the eSafety Commissioner deems to be a public safety threat. The eSafety Commissioner would give a blocking request in written notice to the internet service provider where material is published that threatens public safety. I propose fines for non-complying companies of up to approximately $10 million per offence, similar to European Union penalties. To adequately address this, the fines must be severe and immediate to deter technology companies and developers (Lauren C.Hall et al, 2026). The consequence for individuals in non-compliance is proposed to be a fine of up to $100,000, depending on the severity of harm. This is aligned with the Australian penalty for posting sexually exploitative deepfakes (eSafety Commissioner, 2025). 

Interpretation of Terms

The broadness of the term “material” would be inclusive of AI manipulated and generated content, not limited to deepfake content. This broadness would provide the eSafety Commissioner powers if other AI content threatened public safety, not limiting the powers. 

Truth

In the context of Australian defamation law, the legal definition of “substantial truth” implies the statement must reflect the truth, but not necessarily be completely accurate (Wayne Davis, 2024). This was reflected in the case of Wagner & Ors v Harbour Radio Pty Ltd & Ors in the Queensland Supreme Court where Judge Flanagan said errors in detail are tolerated, when every part of the material truth relied upon in the argument is established (Wayne Davis, 2024). 

Internet Service Provider

This term must be narrow enough to perform within the intentions of the act, but not be too broad as to create oppressive obligations on technology organisations. This should have the meaning designated in the Online Safety Act s19. 

Threaten

This definition should be construed within the definition provided in the Online Safety Act s5. 

Barriers and Risks

Barriers

This policy may be prevented from being implemented by the cost of legislative amendments and extra staff and resources required for the eSafety Commissioner to utilise greater powers (Juan Ornelas et al, 2024). Additionally, greater power to remove AI deepfakes may encounter skepticism as the National AI Plan has a focus of facilitating AI development, which stricter regulations may not align with (Department of Industry, Science and Resources, 2025). 

Risks

There is ambiguity over what threat to public safety qualifies as substantial enough to invoke the eSafety Commissioners powers to enforce the removal of deepfake content by the internet service provider. There are concerns that this will hinder development of AI by increasing regulation and stoking fears of greater liability (Peter Goettler, 2025). This would negatively impact the Australian economy as discouragement of AI development could adversely hinder Australian technological advancement in this area due to fears of legal implications. This could be mitigated with carefully constructed legislation which doesn’t restrict AI development and defines the issue narrowly.

An issue in implementing this proposal is the sufficiency of incentives for internet service providers to comply with requests of the eSafety Commissioner. The providers must act promptly, but it is challenging for this to be enforced. This emergency power of the eSafety Commissioner proposed does not come with penalties for providers that do not comply, thus compliance isn’t adequately incentivised. To combat this, further legislative amendments must be made and fines would potentially need to be issued based on the severity of the breach and inaction.

References

ABC News. (2025, December 1). AI guidance to label and watermark AI content. https://www.abc.net.au/news/2025-12-01/ai-guidance-label-watermark-ai-content/106083786

ABC News. (2025, December 2). National artificial intelligence plan says existing laws are enough. https://www.abc.net.au/news/2025-12-02/national-artificial-intelligence-plan-growth-existing-laws/106086474

AI Act Explorer. AI Act Explorer. https://artificialintelligenceact.eu/ai-act-explorer/

AI Incident Database. AI Incident Database. https://incidentdatabase.ai

Alternative Law Journal. (2025). [Article title unavailable from URL]. https://www.tandfonline.com/doi/full/10.1080/10383441.2025.2504791

Artificial Intelligence Act. Artificial Intelligence Act. https://artificialintelligenceact.eu/

Associated Press. (2025). Denmark eyes new law to protect citizens from AI deepfakes. https://www.ap.org/news-highlights/spotlights/2025/denmark-eyes-new-law-to-protect-citizens-from-ai-deepfakes/

Australian Communications and Media Authority. (2026). Indicator 6: Trust and impact. https://www.acma.gov.au/publications/2026-05/report/indicator-6-trust-and-impact

Australian Cyber Security Centre. Content credentials: Strengthening multimedia integrity in the generative AI era. https://www.cyber.gov.au/business-government/secure-design/artificial-intelligence/content-credentials-strengthening-multimedia-integrity-in-the-generative-ai-era

Australian Government Department of Finance. (2025). Introducing the APS AI Plan. https://www.finance.gov.au/about-us/news/2025/introducing-aps-ai-plan

Australian Government, Office of Impact Analysis. (2024, September 12). Online misinformation and disinformation reform. https://oia.pmc.gov.au/published-impact-analyses-and-reports/online-misinformation-and-disinformation-reform

Australian Human Rights Commission. Select committee on adopting artificial intelligence. https://humanrights.gov.au/__data/assets/file/0021/47244/Select_committee_on_adopting_artificial_intelligence.pdf

Australian National University National Security College. AI, digital literacy and the future of democracy. https://nsc.anu.edu.au/research/ai-digital-literacy-and-future-democracy

Australian Securities and Investments Commission. (2025). We’re not there yet: Current regulation around AI may not be sufficient. https://www.asic.gov.au/about-asic/news-centre/speeches/we-re-not-there-yet-current-regulation-around-ai-may-not-be-sufficient/

BBC News. (2022). Ukraine conflict: How are AI deepfakes being used? https://www.bbc.com/news/technology-60780142

Bell Gully. (2024). Government releases guidance for the responsible use and adoption of generative AI by the public sector. https://www.bellgully.com/insights/government-releases-guidance-for-the-responsible-use-and-adoption-of-generative-ai-by-the-public-sector/

Brookings Institution. (2018). How to combat fake news and disinformation. https://www.brookings.edu/articles/how-to-combat-fake-news-and-disinformation/

Brookings Institution. (2025). Journalism needs better representation to counter AI. https://www.brookings.edu/articles/journalism-needs-better-representation-to-counter-ai/

Canadian Law Review. (2020). [Article title unavailable from URL]. https://www.austlii.edu.au/au/journals/CanLawRw/2020/18.pdf

Clayton Utz. (2025). Transparency and explainability in AI systems: Global perspectives and the road to regulation in Australia. https://www.claytonutz.com/insights/2025/november/transparency-and-explainability-in-ai-systems-global-perspectives-and-the-road-to-regulation-in-australia

Columbia Journal of European Law. (2024). Deepfake, deep trouble: The European AI Act and the fight against AI-generated misinformation. https://cjel.law.columbia.edu/preliminary-reference/2024/deepfake-deep-trouble-the-european-ai-act-and-the-fight-against-ai-generated-misinformation/

Commonwealth Bank of Australia. (2026). Can Australians spot deepfake scams?https://www.commbank.com.au/articles/newsroom/2026/01/can-australians-spot-deepfake-scams.html

Department of Industry, Science and Resources. (n.d.) Being clear about AI-generated content. https://www.industry.gov.au/publications/being-clear-about-ai-generated-content

Department of Industry, Science and Resources. (2025). National AI Plan. https://www.industry.gov.au/publications/national-ai-plan

Department of Industry, Science and Resources. Why we wrote “Being clear about AI-generated content”. https://www.industry.gov.au/publications/being-clear-about-ai-generated-content/why-we-wrote

Digital Transformation Agency. (2025). AI policy update: Strengthening responsible use across government. https://www.dta.gov.au/articles/ai-policy-update-strengthening-responsible-use-across-government

eSafety Commissioner. (2022). Deepfake position statement. https://www.esafety.gov.au/sites/default/files/2022-01/Deepfake-position-statement%20_v2.pdf

eSafety Commissioner. (2025). Court orders $343,500 penalty for posting deepfakes of Australian women. https://www.esafety.gov.au/newsroom/media-releases/court-orders-343500-penalty-for-posting-deepfakes-of-australian-women

Edelman. (2022). 2022 Edelman Trust Barometer: Australia. https://www.edelman.com/au/trust-barometer-2022-australia

European Commission. Apply AI strategy. https://digital-strategy.ec.europa.eu/en/policies/apply-ai

European Commission. (2026, July 31). European approach to artificial intelligence. Shaping Europe’s Digital Future. https://digital-strategy.ec.europa.eu/en/policies/european-approach-artificial-intelligence

European Commission. Regulatory framework proposal on artificial intelligence. https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai

Karttunen, S. (2026, January). The Danish approach to copyright and deepfakes: A model for the EU?. https://www.europarl.europa.eu/RegData/etudes/ATAG/2026/782611/EPRS_ATA(2026)782611_EN.pdf

Garina, N. (2026, June 24). The EU’s AI Transparency Code of Practice, Explained. Tech Policy.Press.  https://www.techpolicy.press/the-eus-ai-transparency-code-of-practice-explained/

Goettler, P. (2025). Why AI overregulation could kill the world’s next tech revolution. https://www.cato.org/commentary/why-ai-overregulation-could-kill-worlds-next-tech-revolution

Information Technology and Innovation Foundation. (2021). How much will the Artificial Intelligence Act cost Europe?https://itif.org/publications/2021/07/26/how-much-will-artificial-intelligence-act-cost-europe/

Institute for Strategic Dialogue. (2024). The EU AI Act: Insights from the world’s first comprehensive AI law. https://www.isdglobal.org/digital_dispatches/the-eu-ai-act-insights-from-the-worlds-first-comprehensive-ai-law/

Kabuz, M. (2024). Deepfakes and the Artificial Intelligence Act - An important signal or a missed opportunity?. https://doi.org/10.1002/poi3.406

Law Society Journal. (2025). Denmark proposes copyright laws to protect against deepfakes. https://lsj.com.au/articles/denmark-proposes-copyright-laws-to-protect-against-deepfakes/

Lowy Institute. (2025). Australia bets old laws to manage new AI risks. https://www.lowyinstitute.org/the-interpreter/australia-bets-old-laws-manage-new-ai-risks

Martin, N. (2025). Online safety regulation of deepfake abuse: A case study on Australia's eSafety Commissioner. Griffith Law Review, 34(1), 23–46. https://doi.org/10.1080/10383441.2025.2504791

Meier, K., & Spichiger, R. (2024, March 15). The EU AI Act: What it means for your business. EY. https://www.ey.com/en_ch/insights/forensic-integrity-services/the-eu-ai-act-what-it-means-for-your-business

MinterEllison. (2025). Australia introduces a national AI plan: Four things leaders need to know. https://www.minterellison.com/articles/australia-introduces-a-national-ai-plan-four-things-leaders-need-to-know

MinterEllison. (2025). Taking responsibility: Guardrails for high-risk AI. https://www.minterellison.com/articles/taking-responsibility-guardrails-for-high-risk-ai

New Zealand Government Digital Services. Misinformation and hallucinations. https://www.digital.govt.nz/standards-and-guidance/technology-and-architecture/artificial-intelligence/responsible-ai-guidance-for-the-public-service-genai/genai-foundations/misinformation-hallucinations

NSW Government. (2025). NSW Government strengthens protections against deepfakes and image-based abuse. https://dcj.nsw.gov.au/news-and-media/media-releases/2025/nsw-government-strengthens-protections-against-deepfakes-and-ima.html

Office of Parliamentary Counsel. Drafting. https://www.opc.gov.au/opc-services/drafting

Parliament of Australia. (2024). Combatting misinformation and disinformation bill digest. https://www.aph.gov.au/Parliamentary_Business/Bills_Legislation/bd/bd2324a/24bd081

Parliament of Australia. Combatting Misinformation and Disinformation Bill – Impact analysis. https://www.aph.gov.au/-/media/Senate/committee/Environment_and_Communications/MDI/Combatting_Misinformation_and_Disinformation_Bill_-_Impact_Analysis.pdf

Parliament of Australia. Impacts of AI on democracy. https://www.aph.gov.au/Parliamentary_Business/Committees/Senate/Adopting_Artificial_Intelligence_AI/AdoptingAI/Interim_report/Chapter_2_-_Impacts_of_AI_on_democracy

Parliamentary Education Office. Amendments. https://peo.gov.au/understand-our-parliament/how-parliament-works/bills-and-laws/amendments

Piper Alderman. (2025). Australia now has a national AI plan. Now what? https://piperalderman.com.au/insight/australia-now-has-a-national-ai-plan-now-what/

Reuters Institute for the Study of Journalism. (2025). Generative AI and news report 2025: How people think about AI’s role in journalism and society. https://reutersinstitute.politics.ox.ac.uk/generative-ai-and-news-report-2025-how-people-think-about-ais-role-journalism-and-society

Senate Environment and Communications Legislation Committee. (2024). Chapter 1: Introduction. In Communications Legislation Amendment (Combatting Misinformation and Disinformation) Bill 2024 [Provisions] (pp. 1–12). Parliament of Australia. https://www.aph.gov.au/Parliamentary_Business/Committees/Senate/Environment_and_Communications/MisandDisinfobill/Report/Chapter_1_-_Introduction

Springer Nature. (2025). [Article title unavailable from URL]. AI and Ethics. https://link.springer.com/article/10.1186/s41235-025-00683-w#Sec3

SSRN. (2025). [Paper title unavailable from URL]. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5259359

Stonegate Legal. (n.d.). Defence of justification (truth) in defamation in Australia. https://stonegatelegal.com.au/defence-of-justification-truth-in-defamation-in-australia/

Swansea University. (2025). Fake or the real thing? How AI can make it harder to trust the pictures we see. https://www.swansea.ac.uk/press-office/news-events/news/2025/11/fake-or-the-real-thing-how-ai-can-make-it-harder-to-trust-the-pictures-we-see.php

The Conversation. (2025). Australia’s national plan says existing laws are enough to regulate AI. This is false hope. https://theconversation.com/australias-national-plan-says-existing-laws-are-enough-to-regulate-ai-this-is-false-hope-271725

The Conversation. (2025). Fines alone won’t stop big tech behaving badly: Here’s what might work. https://theconversation.com/fines-alone-wont-stop-big-tech-behaving-badly-heres-what-might-work-276969

The Guardian. (2025, June 17). More Australians get their news via social media than traditional sources for first time, report finds. https://www.theguardian.com/media/2025/jun/17/more-australians-get-their-news-via-social-media-than-traditional-sources-for-first-time-report-finds

United Nations Association of Australia WA Division. (n.d.). UNAA WA. https://www.unaa-wa.org.au

United Nations University. (2023). Artificial intelligence powered disinformation. https://unu.edu/sites/default/files/2023-09/artificial%20Intelligence%20powered%20disinformation.pdf

University of Canberra. (2024). Digital News Report Australia 2024: AI, social media, misinformation and distrust. https://www.canberra.edu.au/about-uc/media/newsroom/2024/june/digital-news-report-australia-2024-ai,-social-media,-misinformation-and-distrust-what-the-data-tells-us-about-the-news-landscape-in-2024

University of Canberra. Digital News Report Australia. https://www.canberra.edu.au/research/centres/nmrc/digital-news-report-australia

9News. (2025). CommBank research: Most Australians are unable to spot AI-generated scams. https://www.9news.com.au/national/commbank-research-most-australians-are-unable-to-spot-ai-generated-scams/11d952b9-0a57-454d-9b06-c29da34316f8

The views and opinions expressed by Global Voices Fellows do not necessarily reflect those of the organisation or its staff.