Giggle for Girls Pty Ltd v Tickle [2026] FCAFC 64: AI Facial Recognition, Gender Identity and Discrimination
Case Details
| Particular | Details |
|---|---|
| Case | Giggle for Girls Pty Ltd v Tickle [2026] FCAFC 64 |
| Court | Full Court of the Federal Court of Australia |
| Judgement Date | 15 May 2026 |
| Bench | Perry, Abraham and Kennett JJ |
| Legislation | Sex Discrimination Act 1984 (Cth) |
Introduction: When Technology Decides Who Belongs
Artificial intelligence is increasingly being used to decide who gets access, who is rejected and who receives further scrutiny. Facial-recognition systems can identify or classify people; algorithms can assess voices, behaviour and images; automated systems can influence decisions in employment, finance, security and online services.
But what happens when an algorithmic classification affects a person’s legally protected identity?
That question lies at the heart of Giggle for Girls Pty Ltd v Tickle [2026] FCAFC 64, one of the most significant recent Australian judgements at the intersection of AI facial recognition, gender identity and discrimination.
The case concerned Giggle, a women-only social networking application. Users seeking access were required to upload a selfie. Artificial intelligence software assessed the photograph for facial characteristics associated with men and women. Roxanne Tickle, a transgender woman, was initially admitted to the application but was subsequently excluded after her photograph was reviewed by a human decision-maker.
The Full Court unanimously dismissed Giggle’s appeal and allowed Ms Tickle’s cross-appeal. Importantly, the Court did not merely affirm the first-instance judgement. It changed the legal characterisation of the discrimination from indirect discrimination to direct discrimination and increased the compensation from A$10,000 to A$20,000.
The judgement therefore deserves attention not simply as a transgender-rights case, but as an important illustration of how existing anti-discrimination principles can operate when human beings and artificial intelligence jointly participate in deciding who is admitted to a digital service.
Also Read: Abdel-Hady v Commonwealth of Australia [2026]: Unlawful Detention, Rule of Law & Lessons for India
Latest Development: The High Court Refused Special Leave
There is an important development which must be included in any current discussion of the case.
On 10 September 2026, the High Court of Australia refused special leave to appeal in Giggle for Girls Pty Ltd (ACN 632 152 017) & Anor v Tickle [2026] HCADisp 209, with costs. The Federal Court’s official record also records the High Court outcome as “special leave refused with costs”.
This does not mean that the High Court heard a full appeal and affirmed every aspect of the Full Court’s reasoning. Special leave was simply refused. Nevertheless, the Full Court’s judgement remains the operative appellate authority.
The High Court’s refusal is particularly significant because its disposition stated that there was no reason to doubt the Full Court’s construction of s 7D of the Sex Discrimination Act and that an appeal dependent upon establishing error in that construction did not have sufficient prospects of success to warrant special leave.
The Facts: Giggle, Facial Recognition and Roxanne Tickle
Giggle was conceived as a digital women-only space where women could communicate without the presence of men. Registration required a mobile telephone number and a photograph or “selfie”. The photograph was assessed by third-party artificial intelligence software designed to distinguish facial appearances associated with men and women.
The evidence showed that the AI software had deliberately been configured below its maximum reliability. The stated approach was to favour inclusion of someone identifying as a woman rather than exclude a user merely because the AI classified that person as male.
Ms Tickle registered for the Giggle App in approximately February 2021. The AI system initially permitted her access. Her access was later restricted after an individual reviewed her photograph. She sought to regain access, but Giggle refused to restore it.
Ms Tickle subsequently complained to the Australian Human Rights Commission. That complaint was terminated after the Commission’s delegate concluded that there was no reasonable prospect of settlement through conciliation. She then commenced proceedings in the Federal Court.
By August 2022, Giggle had ceased making the application available to anyone. That fact is important because the litigation ultimately concerned conduct that had already occurred rather than an ongoing operating service.
What the First-Instance Court Decided
At first instance, Bromwich J delivered judgement in Tickle v Giggle for Girls Pty Ltd (No 2) [2024] FCA 960.
The primary judge found that Giggle and its founder and CEO, Sally Grover, had engaged in unlawful indirect discrimination against Ms Tickle on the ground of gender identity, contrary to s 22 of the Sex Discrimination Act 1984 (Cth).
The Court awarded Ms Tickle A$10,000 in damages.
The first-instance reasoning was important because the Court accepted that the operation of the app imposed a condition or practice that disadvantaged transgender women. However, the primary judge approached the claim as indirect discrimination rather than direct discrimination.
Giggle and Ms Grover appealed. Ms Tickle filed a cross-appeal seeking, among other things, a finding of direct discrimination and a higher damages award.
The Appeal Outcome: Precisely What the Full Court Decided
The appellate outcome should be stated with precision because two separate appellate proceedings were involved.
| Proceeding | Outcome |
|---|---|
| Giggle and Ms Grover’s appeal | Dismissed |
| Ms Tickle’s cross-appeal | Allowed |
| Original finding of indirect discrimination | Set aside |
| Substituted finding | Direct discrimination |
| Original damages | A$10,000 |
| New damages | A$20,000 |
The Full Court expressly ordered that the declaration of unlawful indirect discrimination made by the primary judge be set aside.
In its place, the Court declared that Giggle and Ms Grover had engaged in unlawful direct discrimination against Ms Tickle on the ground of her gender identity, contrary to s 22 read with s 5B(1) of the Act.
The substituted declaration identified three connected aspects of the unlawful conduct:
- excluding Ms Tickle from the Giggle App on the basis of her gender-related appearance;
- refusing to restore her access on the basis of her gender-related appearance; and
- thereby treating Ms Tickle, a transgender woman, less favourably than a person designated female at birth seeking access to the application.
These formal orders are critical because they demonstrate that the Full Court did not merely uphold the result reached below. It changed the legal basis of liability.
Why the Finding of Direct Discrimination Matters
The distinction between direct and indirect discrimination is not merely semantic.
Direct discrimination focuses upon whether a person was treated less favourably because of a protected characteristic or a characteristic associated with that protected ground.
Indirect discrimination generally concerns a condition, requirement or practice that appears neutral but disadvantages people possessing a particular protected characteristic.
The Full Court concluded that the evidence established direct discrimination under s 5B(1).
That was a significant appellate development because the primary judge had considered the case through the indirect-discrimination framework.
The Full Court therefore moved the legal analysis from:
“Did the platform impose a condition that disadvantaged transgender women?”
to the more direct question:
“Was Ms Tickle treated less favourably because of her gender identity or a characteristic associated with that gender identity?”
Gender Identity and Gender-Related Appearance
The statutory definition of gender identity was central to the appeal.
The Sex Discrimination Act defines gender identity in terms that include gender-related identity and gender-related characteristics, including appearance.
Justice Perry emphasised that the statutory definition controls the meaning of the term within the Act. The Court was therefore not free to substitute a narrower, purely subjective conception of gender identity for the definition Parliament had enacted.
This becomes particularly important when facial recognition is involved.
A facial-recognition system does not ordinarily observe a person’s private identity directly. It assesses physical characteristics and produces a classification based upon them.
If the relevant appearance becomes the reason for exclusion, the fact that the organisation may not know how the person privately identifies does not necessarily end the discrimination inquiry.
Was Actual Knowledge of Transgender Status Necessary?
This is arguably the most important doctrinal point in the judgement.
The Full Court rejected the proposition that actual knowledge that Ms Tickle was transgender was necessarily required for the direct-discrimination claim.
The statutory provisions encompass not only gender identity itself but also relevant characteristics associated with, or imputed to, persons sharing that gender identity.
Consequently, an organisation cannot necessarily defeat a direct-discrimination claim merely by saying:
“We did not know that this person was transgender.”
The legally relevant question may instead concern the characteristic that actually resulted in the less favourable treatment.
This is highly significant for AI governance. Modern AI systems increasingly infer characteristics rather than receive them directly from individuals.
An algorithm may never be told that a person is transgender, belongs to a particular ethnic group or has a particular characteristic. It may nevertheless classify the person using observable or inferred proxies.
Giggle therefore raises a broader principle: lack of express knowledge of a protected identity does not necessarily make discriminatory treatment legally neutral.
The Comparator: Why the Court Focused on a Person Designated Female at Birth
The Full Court’s approach to the comparator also deserves attention.
Giggle sought to frame the comparison in a way that treated the relevant distinction as one between Ms Tickle and a male person.
The Full Court instead considered whether Ms Tickle had been treated less favourably than a person designated female at birth seeking access to the application.
This is important because the comparator must allow the Court to identify the actual source of the less favourable treatment.
If a transgender woman is excluded because her appearance does not conform to the platform’s expectations of female appearance, comparing her simply with a man risks obscuring the real issue.
The Full Court’s substituted declaration expressly used the comparison with a person designated female at birth.
The judgement should not, however, be overstated as resolving every philosophical or legal question concerning the meaning of biological sex. The Court was deciding the statutory discrimination claim before it, on its particular facts and statutory framework.
Also Read: A $150 Million Indigenous Land Judgement: What India Can Learn About Tribal Land Rights
The Section 7D Special-Measures Argument
Giggle argued that its women-only model was a special measure under s 7D of the Sex Discrimination Act, intended to achieve substantive equality between men and women.
Section 7D recognises certain special measures designed to achieve substantive equality.
The Full Court rejected Giggle’s reliance on that provision in the circumstances of this case.
The Court accepted a distributive construction of s 7D. In practical terms, a special measure directed towards achieving substantive equality between one protected group does not automatically provide a defence to discrimination against another protected group.
Thus, even assuming that a women-only service may pursue a legitimate equality objective, that does not itself create a general exemption from the Act’s prohibition against gender identity discrimination.
This point is particularly important because the judgement did not simply declare that every women-only service is unlawful. It decided that the particular statutory special-measure defence advanced by Giggle did not protect the conduct established on the evidence.
The High Court’s later refusal of special leave is also significant here. Its 10 September 2026 disposition stated that there was no reason to doubt the Full Court’s construction of s 7D(1) and (2).
Damages: From A$10,000 to A$20,000
The Full Court also increased the compensation substantially.
The primary judge had awarded A$10,000.
The Full Court substituted an award of A$20,000, comprising:
- A$12,000 general damages; and
- A$8,000 aggravated damages.
The aggravated component is significant because the Court considered aspects of the conduct of the proceeding and public commentary to warrant additional compensation. The Court did not treat the appellants’ asserted bona fide beliefs as sufficient to eliminate the basis for aggravated damages.
The Court also ordered Giggle and Ms Grover to pay Ms Tickle’s costs of the appeal and cross-appeal, subject to the relevant A$50,000 caps.
An Important Point: What the Full Court Did Not Decide
A careful legal article must also identify what the court did not decide.
The litigation had raised constitutional questions concerning, among other things, Commonwealth legislative power.
However, the constitutional grounds were ultimately withdrawn on appeal. Consequently, the Full Court did not deliver a final appellate ruling resolving all of those constitutional questions.
It would therefore be inaccurate to describe Giggle as a Full Court determination conclusively establishing the constitutional validity of every relevant gender-identity provision.
The strength of the decision lies principally in its statutory interpretation and application of the Sex Discrimination Act.
The Real Technology Lesson: AI Does Not Become a Legal Shield
One of the most important lessons is that this was not a case of a completely autonomous machine making an unreviewed decision.
The sequence was more complicated:
- Ms Tickle supplied a photograph;
- AI software assessed the image;
- the system initially granted access;
- a human subsequently reviewed the photograph;
- access was restricted;
- Ms Tickle sought reinstatement; and
- Giggle refused to restore access.
This is the modern human-machine decision chain.
It is increasingly common in employment, banking, insurance, online platforms, border control, policing and government administration.
The legal responsibility therefore cannot be analysed simply by asking whether “the AI made the decision”.
The better questions are:
- Who designed the system?
- Who deployed it?
- What classification did it produce?
- How was that classification used?
- Who reviewed it?
- Could the reviewer depart from it?
- Was the person given an opportunity to challenge the outcome?
- Who ultimately authorised the adverse decision?
The central lesson of AI facial recognition, gender identity and discrimination is therefore straightforward:
Technology may change the mechanism of decision-making, but it does not automatically change the legal responsibility for the outcome.
Human Review Must Be Meaningful
The phrase “human in the loop” is increasingly used in AI governance.
But a human reviewer does not automatically make a decision lawful.
If the human merely confirms an algorithm’s conclusion, the human review may have little practical value.
Effective human oversight should allow the decision-maker to understand the limitations of the system, examine relevant evidence independently and overturn an automated classification where appropriate.
Giggle therefore provides a useful warning for technology companies: human oversight must be substantive, not merely procedural.
The Larger Principle: Algorithmic Discrimination
Facial recognition may be only the beginning.
AI systems increasingly classify individuals using:
- faces and photographs;
- voice patterns;
- language;
- behaviour;
- location;
- purchasing patterns;
- biometric information;
- employment history; and
- online activity.
Some systems may infer characteristics that an individual has never expressly disclosed.
This creates a serious legal problem. If a protected characteristic is inferred from a proxy, should an organisation be able to escape liability simply because the individual never declared that characteristic?
Giggle suggests that the answer cannot automatically be yes.
The future of discrimination law may therefore increasingly involve a shift from asking whether a human decision-maker consciously intended to discriminate to examining what information actually drove the adverse treatment.
What India Can Learn From Giggle for Girls v Tickle
The Australian judgement deserves particular attention in India.
India is rapidly expanding the use of artificial intelligence, biometric authentication, facial-recognition technologies, CCTV analytics and automated verification. Aadhaar and other digital systems have also made biometric identity a major part of India’s technological infrastructure.
India’s Digital Personal Data Protection framework is an important development. But privacy and anti-discrimination are different legal questions.
A system may lawfully process personal data and yet produce a discriminatory outcome.
India should therefore consider a stronger framework for high-impact AI systems, including:
- Algorithmic impact assessments before deployment in high-risk environments;
- independent bias and accuracy testing;
- greater transparency concerning consequential automated decisions;
- meaningful human review;
- accessible appeal and correction mechanisms;
- heightened safeguards for biometric profiling; and
- clear legal responsibility for organisations deploying AI systems.
The constitutional values embodied in Articles 14 and 21 of the Indian Constitution—particularly equality, non-arbitrariness, dignity and privacy—provide an important foundation for this discussion.
Why This Matters for Aadhaar, CCTV and Facial Recognition in India
The Indian debate should not be reduced to whether facial recognition should be permitted or prohibited.
The more useful legal question is: under what circumstances should facial recognition be permitted, what safeguards should accompany it, and what remedy should an individual have when the system is wrong?
Consider a hypothetical system that incorrectly classifies a person and consequently denies access to a service.
If the organisation responds, “The computer rejected you,” that should not end the legal inquiry.
A constitutional democracy cannot outsource accountability to software.
The same principle becomes even more important when the technology is used by public authorities or in circumstances affecting fundamental rights.
Critical Legal Analysis
From the perspective of a practising lawyer, the most enduring significance of Giggle for Girls v Tickle lies in the Court’s refusal to allow technological complexity to obscure the ordinary operation of statutory rights.
An algorithm may produce a score.
A facial-recognition system may produce a classification.
A company may describe the classification as objective.
But the court must still ask:
- What characteristic actually influenced the decision?
- Was the person treated less favourably?
- Does the characteristic fall within a protected statutory ground?
- Was the conduct authorised by a valid exception?
- Was the human review genuine?
- Was there a meaningful opportunity to challenge the result?
These remain questions of law, evidence and statutory interpretation.
Technology may complicate the evidence. It does not eliminate the law.
Global Significance: The Law Is Moving Into the Age of Automated Identity
The deeper significance of this case extends beyond Australia.
For much of the twentieth century, discrimination law assumed that one human being was making a decision about another human being.
That assumption is becoming increasingly outdated.
Today, decisions can be influenced by machine-learning models, facial-recognition software, predictive scoring systems and automated classification tools.
The emerging legal challenge is therefore not simply whether AI is accurate.
It is whether AI is lawful, fair, explainable and accountable when its classifications affect human rights.
That is why Giggle for Girls v Tickle deserves international attention.
Conclusion: Algorithms Must Remain Subject to the Law
Giggle for Girls Pty Ltd v Tickle [2026] FCAFC 64 is an important judgement at the intersection of AI facial recognition, gender identity and discrimination.
The Full Court did not simply uphold the original judgement. It dismissed Giggle’s appeal, allowed Ms Tickle’s cross-appeal, replaced the finding of indirect discrimination with a finding of direct discrimination, and increased compensation from A$10,000 to A$20,000.
The Court specifically found that exclusion and refusal to restore access occurred on the basis of Ms Tickle’s gender-related appearance.
The Court also rejected Giggle’s reliance upon the special-measures provision in s 7D in the circumstances of the case. The subsequent refusal of special leave by the High Court on 10 September 2026 leaves the Full Court judgement standing as the operative appellate authority.
For India, the lesson is particularly timely.
As artificial intelligence, biometric systems, facial recognition and automated decision-making become increasingly embedded in public and private life, legal accountability must develop alongside technological capability.
The central principle should be simple:
An algorithm cannot become a legal shield for a discriminatory decision merely because a machine participated in making it.
The more profound question for the coming decade is therefore not merely what AI can recognise, but what legal consequences should follow when AI decides that a person belongs—or does not belong—to a particular category?
That is the enduring significance of Giggle for Girls v Tickle.
Case Citation and Primary Authorities
| Authority | Details |
|---|---|
| Giggle for Girls Pty Ltd v Tickle [2026] FCAFC 64 | Full Court of the Federal Court of Australia, 15 May 2026. |
| Tickle v Giggle for Girls Pty Ltd (No 2) [2024] FCA 960 | Federal Court of Australia, 23 August 2024. |
| Giggle for Girls Pty Ltd (ACN 632 152 017) & Anor v Tickle [2026] HCADisp 209 | High Court of Australia, 10 September 2026 — special leave refused with costs. |
| Sex Discrimination Act 1984 (Cth) | Particularly ss 4, 5B, 7D and 22. |
Facing an AI, Facial Recognition or Discrimination Issue? Take Legal Advice Now.
Have you been affected by an AI-driven decision, facial-recognition system, biometric profiling, automated exclusion or discriminatory digital platform?
Do you need to understand your legal rights, assess potential discrimination, challenge an automated decision or determine what remedies may be available under Indian law?
Do not leave a technology-driven decision unchallenged simply because “the algorithm decided it”. Get your legal position assessed by an experienced Supreme Court advocate.
Adv. Tarun Choudhury
Supreme Court Advocate | 25+ Years of Legal Experience
📞 Call now: 9650499965
💬 WhatsApp now: 8851978611
Take the first step. Contact Adv. Tarun Choudhury for legal guidance on AI, facial recognition, biometric profiling, algorithmic discrimination, privacy, constitutional rights and emerging technology-law disputes.
Frequently Asked Questions About Giggle for Girls v Tickle
1. What is the significance of Giggle for Girls Pty Ltd v Tickle [2026] FCAFC 64?
Giggle for Girls Pty Ltd v Tickle [2026] FCAFC 64 is a significant Australian judgement concerning AI facial recognition, gender identity and discrimination.
The Full Court of the Federal Court of Australia dismissed Giggle’s appeal and allowed Roxanne Tickle’s cross-appeal. The Court substituted a finding of direct discrimination for the first-instance finding of indirect discrimination and increased compensation from A$10,000 to A$20,000.
The case demonstrates that the use of AI or facial-recognition technology does not automatically protect a platform from anti-discrimination law.
2. Can facial-recognition technology result in unlawful discrimination?
Yes, potentially. Facial-recognition technology can create discriminatory outcomes when a person’s appearance or another protected characteristic is used as the basis for less favourable treatment.
In Giggle for Girls v Tickle, the Full Court found direct discrimination on the ground of gender identity after Ms Tickle was excluded from the women-only app based on her gender-related appearance.
The judgement illustrates why AI-based classification should remain subject to human rights and anti-discrimination safeguards.
3. Did the Australian Court require Giggle to know that Roxanne Tickle was transgender?
No. The Full Court held that actual knowledge of Ms Tickle’s transgender status was not a necessary element of the relevant direct-discrimination claim.
The statutory concept of gender identity under Australia’s Sex Discrimination Act 1984 includes relevant gender-related characteristics and appearance.
Therefore, the absence of express knowledge of a person’s transgender identity does not necessarily prevent a finding of direct discrimination where the relevant gender-related characteristic influenced the treatment.
4. What can India learn from Giggle for Girls v Tickle about AI discrimination?
India can draw important lessons concerning AI governance, facial recognition, biometric profiling and algorithmic discrimination.
High-impact AI systems should be subject to appropriate accuracy and bias testing, transparency, meaningful human review, accountability and effective mechanisms for challenging erroneous automated decisions.
India’s constitutional principles of equality, non-arbitrariness, privacy and dignity provide an important framework for examining discriminatory AI systems, particularly where automated decisions affect access to important services or rights.
5. Can a company avoid legal liability by saying that an AI algorithm made the decision?
No, the use of an AI algorithm does not automatically provide a legal defence against discrimination.
The important question is how the technology was used and who ultimately relied upon or implemented its output.
The Giggle case involved both AI-assisted facial classification and subsequent human decision-making. It demonstrates the importance of examining the entire human-machine decision chain.
Businesses deploying AI should therefore have appropriate safeguards, meaningful human oversight and effective review mechanisms to identify and correct potentially discriminatory outcomes.
Key Takeaways: Giggle for Girls v Tickle [2026] FCAFC 64
- Giggle for Girls v Tickle [2026] FCAFC 64 is a significant Australian judgement at the intersection of AI facial recognition, gender identity and discrimination.
- The Full Court of the Federal Court of Australia dismissed Giggle’s appeal and allowed Roxanne Tickle’s cross-appeal, changing the finding from indirect discrimination to direct discrimination.
- The Court held that excluding Ms Tickle from the women-only digital platform based on her gender-related appearance constituted unlawful direct discrimination under the Sex Discrimination Act 1984 (Cth).
- Actual knowledge of a person’s transgender status was not necessarily required for the relevant direct-discrimination finding where treatment was based on a characteristic associated with gender identity.
- The judgement demonstrates that AI facial-recognition technology does not operate outside anti-discrimination law. Technology cannot automatically shield an organisation from legal responsibility.
- The case involved a human-machine decision chain: AI-assisted facial assessment was followed by human review and exclusion. Human involvement does not automatically cure a discriminatory outcome.
- The Full Court rejected Giggle’s reliance on the special-measures defence under s 7D in the circumstances of the case. A special measure directed towards one protected ground does not automatically authorise discrimination on another protected ground.
- The Full Court increased compensation from A$10,000 to A$20,000, including general and aggravated damages.
- On 10 September 2026, the High Court of Australia refused special leave to appeal, leaving the Full Court’s judgement as the operative appellate authority.
- For India, the judgement highlights the need for stronger safeguards concerning AI discrimination, facial recognition, biometric profiling, algorithmic accountability, human oversight and automated decision-making.
- India can consider stronger governance mechanisms for high-impact AI, including algorithmic impact assessments, bias testing, transparency, meaningful human review and accessible appeal mechanisms.
- The broader legal principle is clear: an algorithm cannot become a legal shield for discriminatory treatment merely because a machine participated in the decision-making process.
