Fact-checked by Grok 2 weeks ago

Berne three-step test

The Berne three-step test is a foundational limitation on exceptions to copyright protection under international law, enshrined in Article 9(2) of the Berne Convention for the Protection of Literary and Artistic Works (Paris Act of 24 July 1971), which permits member states to authorize reproduction of protected works without the author's consent solely in certain special cases, provided such reproduction does not conflict with the normal exploitation of the work and does not unreasonably prejudice the legitimate interests of the author.^[1] Originating from the 1967 Stockholm Diplomatic Conference revising the Berne Convention, the test emerged as a compromise to accommodate emerging technologies like photocopying while safeguarding authors' economic rights amid debates over unrestricted reproduction.^[2] Initially confined to the reproduction right, its principles were generalized in subsequent treaties, including Article 13 of the WTO Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS), extending the criteria to exceptions for all exclusive copyright rights, and influencing digital-era instruments like the WIPO Copyright Treaty.^[3] The test's application has shaped national doctrines, such as fair use in the United States, with WTO dispute settlement panels—most notably in the 2000 ruling on U.S. Copyright Act Section 110(5)—interpreting its terms to allow interpretive flexibility while rejecting broad categorical exemptions that fail the cumulative steps, thereby balancing public access against rights holders' incentives.^[4] Debates persist over its scope, with some analyses advocating narrower readings to protect creators and others emphasizing empirical evidence of minimal prejudice in limited uses to promote innovation and education, though judicial and scholarly consensus underscores its role in preventing systemic erosion of copyright's core exclusivity.^[5]^[6]

Historical Development

Inception at the 1967 Stockholm Conference

The 1967 Diplomatic Conference in Stockholm, convened from July 11 to 24 under the auspices of the Berne Union for the Protection of Literary and Artistic Works, aimed to revise the Convention to explicitly codify the reproduction right in response to emerging reprographic technologies, particularly photocopying machines, which posed significant risks to authors' economic interests by enabling widespread unauthorized copying.^[7] Prior to this revision, the Berne Convention lacked a comprehensive reproduction right, relying instead on implied protections, but the proliferation of photocopiers in libraries, educational institutions, and businesses highlighted the need for a formal grant of exclusive control over reproduction while anticipating potential exceptions for socially beneficial uses.^[8] The conference addressed these tensions by introducing Article 9(1), affirming authors' exclusive right to authorize reproductions in any manner or form, thereby aligning international standards with national laws increasingly recognizing such rights.^[9] Preparatory efforts by Study Group IV, focused on reprographic reproduction, proposed mechanisms to permit limited exceptions to this new right, recognizing that absolute exclusivity could hinder legitimate activities such as private study, research, or incidental copying.^[8] The United Kingdom delegation advanced a pivotal draft confining exceptions to "certain special cases" that would not conflict with the normal exploitation of the work nor unreasonably prejudice the author's legitimate interests, providing a flexible yet bounded standard rather than an exhaustive enumeration of permitted uses.^[8] This approach contrasted with proposals from civil law jurisdictions favoring enumerated exceptions or compulsory licensing schemes, as suggested by delegations from India and Romania, which sought broader accommodations for developing economies and educational needs.^[10] Conference debates underscored the imperative to safeguard authors' incentives for creation against expansive exceptions that could erode market value, with representatives from authors' and publishers' organizations arguing vigorously for stringent limitations to prevent systemic undermining of reproduction rights.^[8] The resulting compromise embedded this tripartite framework in Article 9(2) of the Stockholm Act, establishing it as a mandatory criterion for any national exceptions to reproduction, thereby balancing proprietary interests with calibrated public access without enumerating specific scenarios.^[9] This formulation reflected a consensus that exceptions should remain exceptional and non-disruptive to core economic exploitation, informed by empirical concerns over photocopying's scale rather than abstract policy preferences.^[7]

Adoption in the 1971 Paris Act of the Berne Convention

The three-step test was formally codified as Article 9(2) in the Paris Act of the Berne Convention, signed on July 24, 1971, following its initial draft in the 1967 Stockholm Revision. This integration established the test as the primary limitation on exceptions to the newly affirmed exclusive reproduction right under Article 9(1), which grants authors the right to authorize reproduction of their works in any manner or form. The Paris Act's adoption affirmed minimum standards for Union members, requiring national laws to balance author protections with permissible limitations through the test's criteria.^[11]^[12] Under Article 9(2), exceptions to the reproduction right are confined to certain special cases that neither conflict with a work's normal exploitation nor unreasonably prejudice the author's legitimate interests. The provision applies solely to the reproduction right delineated in Article 9(1), excluding other enumerated rights—such as translation under Article 8 or adaptation under Article 12—unless the Convention text explicitly incorporates the test for those domains. This targeted scope underscores the test's role in delineating boundaries for reproduction-specific exceptions without broader encroachment on core copyright entitlements.^[13]^[11] The 1971 Paris Act's ratification process integrated the three-step test as a non-derogable framework, obligating member states to align their exception regimes accordingly upon accession. As of 2025, 182 countries adhere to the Berne Convention, subjecting their reproduction right exceptions to this test and ensuring its enforcement as a foundational international benchmark.^[12]^[14]

Legal Provisions and Principles

Text of Article 9(2) Berne Convention

Article 9(1) of the Berne Convention for the Protection of Literary and Artistic Works establishes that authors shall have the exclusive right of authorizing the reproduction of their works, in any manner or form.^[15] Article 9(2) qualifies this right by permitting member states to enact legislation allowing limited exceptions, stating verbatim: "It shall be a matter of legislation in the countries of the Union to determine the conditions under which the provisions of the preceding paragraph shall be applied. It shall be a matter of legislation in the countries of the Union to permit the reproduction of such works in certain special cases, provided that such reproduction does not conflict with a normal exploitation of the work and does not unreasonably prejudice the legitimate interests of the author."^[15] The scope of Article 9(2) applies exclusively to exceptions and limitations on the reproduction right under Article 9(1), enabling national laws to define applicable conditions while adhering to the specified criteria of special cases, non-conflicting exploitation, and non-unreasonable prejudice to authors' interests.^[13] This framework does not extend to compulsory licenses, which are governed by other provisions such as Articles 11, 11bis, and 13, nor does it alter the Berne Convention's minimum standards for other exclusive rights.^[16] The wording of Article 9(2) emerged from the draft revision at the 1967 Stockholm Conference of the Berne Convention, where it was formulated to balance reproduction rights with permissible national flexibilities, and was adopted unchanged in substance into the 1971 Paris Act following diplomatic refinements.^[11]^[2]

Detailed Analysis of the Three Steps

The Berne three-step test establishes cumulative criteria that any exception or limitation to the reproduction right must satisfy to preserve the economic incentives underlying copyright protection, ensuring that such measures do not systematically undermine authors' ability to derive revenue from their works.^[17] Each step builds on the previous, requiring exceptions to be narrowly tailored to avoid broad dilutions that could erode market-based rewards for creation, as evidenced by the test's design to limit ad hoc judicial expansions.^[4] This framework prioritizes verifiable impacts on exploitation opportunities over expansive user claims, aligning with the causal link between exclusive rights and investment in original expression.^[18] Step 1: Certain special cases. The first criterion mandates that exceptions apply only in "certain special cases," interpreted as narrowly defined, predetermined scenarios rather than open-ended doctrines that permit case-by-case determinations.^[3] This confines limitations to well-delineated areas, such as quotations or incidental reproductions, to prevent generalized exemptions that could erode the exclusivity needed for market incentives.^[4] Broad interpretations, like those akin to fair use allowing unpredictable applications, fail this step by introducing uncertainty that causally discourages investment in works susceptible to unforeseen uses.^[19] Step 2: No conflict with normal exploitation. Assuming the first step is met, the exception must not conflict with the "normal exploitation" of the work, focusing on actual or potential markets where authors license or directly monetize their creations.^[20] This evaluates whether the use acts as a substitute that displaces revenue streams, such as by bypassing established licensing models for reproduction in education or media, thereby preserving the causal chain from creation to economic return.^[18] Empirical assessment of market harm is key; non-substitutive uses, like private caching without distribution, may pass, but those enabling widespread unauthorized dissemination conflict by reducing demand for authorized versions.^[19] Step 3: No unreasonable prejudice to legitimate interests. Finally, the exception must not cause "unreasonable prejudice to the legitimate interests" of the author, emphasizing economic harms over non-verifiable public interest assertions, with "legitimate interests" encompassing both exploitation rights and moral protections against distortion.^[21] This step permits balancing only if prior conditions hold and prejudice is demonstrably limited, such as through remuneration schemes that offset lost revenue, ensuring exceptions do not disproportionately impair incentives without compensating data on net societal gain.^[22] Subjective claims of overriding public benefit yield to evidence of unreasonable economic detriment, maintaining the test's role in safeguarding authorial investments against unchecked user expansions.^[4]

Incorporation into Broader International Frameworks

Role in the TRIPS Agreement (1994)

The Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS), established in 1994 under the World Trade Organization (WTO) as an outcome of the Uruguay Round negotiations (1986–1994), incorporated a version of the Berne three-step test in Article 13 to regulate limitations and exceptions across copyright exclusive rights.^[16] This provision requires WTO members to limit such exceptions to "certain special cases" that "do not conflict with a normal exploitation of the work" and "do not unreasonably prejudice the legitimate interests of the right holder."^[16] By generalizing the test from Berne Convention Article 9(2)'s focus on reproduction rights alone, Article 13 extended its constraints to all enumerated exclusive rights, including reproduction, distribution, rental, public performance, and adaptation, thereby promoting uniform standards for IP protection in global trade contexts.^[23]^[6] This expansion addressed gaps in the Berne framework by applying the test prophylactically to prevent broad or unchecked exceptions that could undermine commercial viability of works in international markets, where divergent national laws previously hindered cross-border enforcement.^[23] TRIPS Article 13 thus served as a harmonization tool, mandating compliance with Berne minima while subjecting IP exceptions to trade-linked scrutiny, distinct from Berne's administration via voluntary WIPO mechanisms.^[16] The provision's wording closely mirrors Berne but omits the reproduction-specific qualifier, enabling its use to evaluate exceptions holistically rather than right-by-right.^[6] Empirically, Article 13's embedding in TRIPS facilitated enforceable IP norms through the WTO's Dispute Settlement Understanding, where violations could trigger authorized trade countermeasures, such as tariffs or quotas, incentivizing members to align domestic laws with the test's criteria.^[23] Since 1995, over 160 WTO members have ratified TRIPS, leading to widespread legislative adjustments to copyright exceptions for conformity, though implementation varies in stringency.^[23] This mechanism has bolstered causal protections for creators' economic interests by tying IP adherence to broader trade obligations, reducing incentives for exceptions that erode market-based incentives without equivalent public benefits.^[24]

Inclusion in WIPO Internet Treaties (1996)

The WIPO Copyright Treaty (WCT) and WIPO Performances and Phonograms Treaty (WPPT), adopted on December 20, 1996, during diplomatic conferences in Geneva, incorporated the Berne three-step test to address emerging challenges from digital technologies, such as unauthorized online reproduction and distribution that undermined authors' and performers' exclusive rights.^[3] These treaties responded to the internet's facilitation of widespread, low-cost copying, which causally eroded incentives for creative production by bypassing traditional exploitation mechanisms like licensing.^[4] Article 10 of the WCT mandates that Contracting Parties confine limitations or exceptions to exclusive rights to certain special cases that do not conflict with normal exploitation and do not unreasonably prejudice legitimate interests, directly applying the three-step test to rights updated for the digital environment, including distribution and rental rights.^[25] Similarly, Article 16(2) of the WPPT extends the three-step test to limitations on performers' and producers' rights in phonograms, ensuring exceptions in the digital realm—such as caching or transient copies—remain narrowly tailored to avoid undermining economic value.^[26] An Agreed Statement to Article 10 of the WCT clarifies that the test accommodates extensions of pre-digital Berne-compatible exceptions into online contexts and allows new ones suited to network environments, provided they satisfy the criteria, with specific endorsement for temporary reproductions essential to digital transmission as potentially exempt if transient and integral to a technological process.^[3] This framework reinforces Berne Convention minima by subjecting digital-age exceptions to the same evidentiary rigor, preventing overbroad carve-outs that could nullify protections against internet-enabled infringement. As of the latest records, the WCT binds 118 Contracting Parties, while the WPPT has comparably extensive ratification, collectively obligating over 100 nations to apply the test to performative and phonogram rights alongside literary and artistic works, thus harmonizing global standards amid technological disruption.^[27] These provisions establish the three-step test as a baseline for balancing innovation with proprietary control in digital dissemination, without expanding exceptions beyond what empirical assessments of harm to rightholders would justify.^[4]

Variations and Extensions in Subsequent Agreements

The European Union's Directive 2001/29/EC on the harmonisation of certain aspects of copyright and related rights in the information society incorporates the Berne three-step test verbatim in Article 5(5), stipulating that the directive's enumerated exceptions and limitations "shall only be applied in certain special cases which do not conflict with the normal exploitation of the work or other subject-matter and do not unreasonably prejudice the legitimate interests of the right holder."^[28] This formulation maintains the core criteria while embedding them within a closed list of mandatory and optional exceptions, allowing member states limited flexibility to implement additional safeguards, such as for caricature, parody, or pastiche, provided they satisfy the test's constraints.^[28] Subsequent EU instruments, including the 2019 Directive (EU) 2019/790 on copyright in the Digital Single Market, extend the test's application to emerging areas like text and data mining, with Article 3 permitting such uses without rightholder authorization under conditions aligned with the three steps, thereby adapting the framework to digital realities without altering its foundational logic. In bilateral and plurilateral free trade agreements post-1996, the test appears with phrasing that reinforces rightholder protections, particularly in the third step. The United States-Mexico-Canada Agreement (USMCA), effective July 1, 2020, includes Article 20.H.9, which requires parties to "confine limitations or exceptions to exclusive rights to certain special cases that do not conflict with a normal exploitation of the work, and do not unreasonably prejudice the legitimate interests of the right holder," directly echoing Berne while integrating it into enforceable obligations among signatories with varying development levels. Similarly, the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), entering force on December 30, 2018, for initial parties, mandates in Article 18.65 that exceptions comply with the three-step criteria, supplemented by Article 18.66's call for parties to seek "an appropriate balance between the exclusive rights" and exceptions' promotion of dissemination, introducing a normative layer that interprets flexibility through the lens of equilibrium rather than unilateral expansion. These provisions, often advocated by IP-exporting nations like the United States, tend to prioritize minimizing prejudice to legitimate interests in high-enforcement contexts, contrasting with concessions in pacts involving developing economies that preserve space for access-oriented exceptions without diluting the test's universality. Across these agreements, deviations remain marginal—such as contextual applications to specific rights or interpretive balances—preserving the test as a Berne-derived benchmark that constrains rather than expands exceptions, with tighter Step 3 formulations in developed-led negotiations reflecting incentives for robust exploitation amid global digital trade pressures.^[29] No agreements introduce wholly novel steps or abandon the sequential evaluation, underscoring the test's enduring adaptability without erosion of its restrictive intent.

International Interpretations and Applications

WTO Dispute Settlement Panels, e.g., US-EC Section 110(5) Case (2000)

The WTO Dispute Settlement Body (DSB) has interpreted the Berne three-step test, via its incorporation into Article 13 of the TRIPS Agreement, as imposing cumulative and stringent requirements on copyright exceptions in trade-related disputes, ensuring they do not undermine exclusive rights under Berne Articles 11bis(1)(iii) and 11(1)(ii).^[30] Panels apply the test sequentially, requiring an exception to satisfy each condition independently, with failure at any step resulting in inconsistency with TRIPS obligations.^[31] This approach demands evidence-based analysis, particularly for assessing "normal exploitation" and "legitimate interests," often evaluating market data on revenue streams affected by the exception.^[32] In the landmark United States – Section 110(5) of the US Copyright Act (DS160) dispute, initiated by the European Communities on May 26, 1999, the panel report of June 15, 2000, scrutinized two exemptions in Section 110(5) for public performances of non-dramatic musical works via radio or television in commercial settings without royalties.^[30] The "homestyle" exemption under 110(5)(A)—limited to small speakers and receivers in eating/drinking establishments under 3,750 square feet—was deemed a "certain special case" (step 1) due to its narrow, clearly defined scope; it did not conflict with normal exploitation (step 2), as it affected only incidental, non-commercial-like uses; and it did not unreasonably prejudice right holders' legitimate interests (step 3), given its minimal market impact.^[31] Thus, 110(5)(A) complied with Article 13.^[30] Conversely, the broader "business exemption" under 110(5)(B)—covering larger establishments up to 10,000 square feet for certain non-musical programming—failed step 3 despite passing steps 1 and 2.^[31] The panel held it unreasonably prejudiced legitimate interests by depriving composers and publishers of royalties for public performances, a key economic exploitation channel, as evidenced by collective management organizations collecting fees for such uses in normal markets; the exemption's scope exempted a substantial revenue segment without sufficient countervailing justification.^[32] This interpretation emphasized empirical review of affected markets, rejecting blanket deference to national characterizations of "minor" impacts.^[31] The DSB adopted the panel report on June 26, 2000, with the US appealing aspects of Article 13's applicability, but the Appellate Body upheld the findings on July 18, 2000.^[30] The US notified partial implementation on March 23, 2001, but the EC contested full compliance, leading to a mutually agreed solution on January 28, 2002, involving compensation rather than full repeal of 110(5)(B).^[30] This outcome reinforced the test's function in curbing exceptions that mask protectionist measures, promoting harmonized enforcement of creators' rights across WTO members without favoring domestic industries.^[33]

Influence on Global Treaty Negotiations and Compliance

The Berne three-step test has shaped post-TRIPS treaty negotiations by serving as a benchmark for constraining exceptions, particularly in efforts to extend its application beyond reproduction rights to all exclusive copyright entitlements. In the Anti-Counterfeiting Trade Agreement (ACTA), finalized in October 2011 after negotiations spanning 2007–2010, Article 2.16 mandated that limitations or exceptions be confined to special cases not conflicting with normal exploitation of works and not unreasonably prejudicing rightholders' legitimate interests, directly mirroring the test to curb expansive user privileges amid concerns over digital piracy. Similarly, during Trans-Pacific Partnership (TPP) talks from 2008 to 2015, U.S. proposals advocated broadening the test's scope in Article 18.65 to cover all limitations, rejecting looser formulations in favor of rigorous criteria to prevent "exception creep"—the gradual enlargement of carve-outs that could erode creators' market incentives.^[34] This diplomatic insistence, echoed by rightholder groups, ensured the test's integration as a safeguard against ideologically driven expansions, prioritizing structured evaluation over unfettered national discretion.^[35] The test's clarity has incentivized state compliance by linking exception design to demonstrable alignment with economic realities, reducing incentives for overreach. Post-2000, international frameworks have seen minimal formal challenges to exception regimes under the test, attributable to its demand for exceptions to be narrowly tailored and justified by absence of verifiable harm to exploitation or interests, as evidenced in treaty texts and WIPO discussions.^[36] This has promoted evidence-based policymaking, where states must substantiate non-prejudice through market impact assessments rather than abstract public interest claims, fostering uniformity in compliance across diverse jurisdictions.^[5] In bilateral and plurilateral pacts, such as those under the WTO's TRIPS framework, adherence to the test has deterred violations by embedding it as a compliance litmus, with negotiators citing its role in maintaining balance without frequent recourse to enforcement mechanisms.^[23]

National and Judicial Implementations

Compatibility with Fair Use in Common Law Jurisdictions

The United States maintains that its fair use doctrine under 17 U.S.C. § 107 is compatible with the Berne Convention's three-step test in Article 9(2), as asserted during its 1989 accession to the Convention.^[37] This position holds that the doctrine's case-by-case adjudication, guided by four statutory factors—purpose and character of use, nature of the work, amount and substantiality used, and effect on potential market—ensures exceptions remain confined to special cases without conflicting with normal exploitation or unreasonably prejudicing authors' legitimate interests.^[37] Other common law jurisdictions adopting or expanding fair use, such as Israel in 2011 and Singapore in 2006, have similarly aligned their implementations with the test, emphasizing judicial flexibility to meet international obligations.^[38] The alignment operates through interpretive convergence: the first factor (often favoring transformative, non-commercial uses) narrows exceptions to special cases under step one, while the fourth factor directly evaluates market harm, corresponding to steps two and three by safeguarding economic incentives.^[37] Courts apply these factors cumulatively, providing foreseeable legal certainty as required by interpretations of the test, such as in WTO analyses of related exceptions.^[38] This structure has preserved doctrinal harmony, with U.S. fair use jurisprudence evolving since the 1989 Berne entry without necessitating legislative overhaul to satisfy the test.^[37] Tensions arise from fair use's open-ended nature, potentially risking non-compliance with step one's "special cases" requirement absent explicit enumeration, as critics argue it could enable overly broad applications.^[37] However, caselaw constraints—such as emphasizing transformative use in decisions like Campbell v. Acuff-Rose Music, Inc. (1994)—mitigate this by ensuring exceptions are not general but context-specific, aligning empirically with the test's limits.^[37] Post-2000 developments, including adoptions in other jurisdictions and absence of successful international challenges to fair use itself, demonstrate minimized conflicts, upholding incentives for creators while accommodating public interests.^[38]

Applications in Civil Law Systems and Specific Court Rulings

In civil law jurisdictions, particularly within the European Union, the Berne three-step test has been integrated into national copyright frameworks via the EU InfoSoc Directive (2001/29/EC), which mandates that enumerated exceptions comply with its criteria under Article 5(5). Courts in these systems, emphasizing codified limitations rather than open-ended doctrines, rigorously apply the test to assess whether specific exceptions—such as those for temporary reproductions or private use—meet the requirements of special cases, non-conflict with normal exploitation, and no unreasonable prejudice to legitimate interests. This application often results in narrow interpretations, prioritizing evidence of actual economic impact over presumptive user benefits. The Court of Justice of the European Union (CJEU) has invoked the three-step test in landmark rulings to constrain exceptions. In Infopaq International A/S v Danske Dagblades Forening (Case C-5/08, July 16, 2009), the CJEU examined whether scanning newspaper articles for automated keyword searches qualified under the temporary reproduction exception of Article 5(1)(a). The court ruled that such acts must satisfy all three steps cumulatively: they apply only in special cases (e.g., transient and integral to a technological process), do not conflict with normal exploitation (assessed via potential market substitution), and avoid unreasonable prejudice (requiring empirical justification of harm). The exception was upheld narrowly for purely temporary copies but rejected for the scanning process itself, as it involved deliberate reproduction exceeding technical necessity.^[39]^[29] German courts have similarly scrutinized private copying exceptions, which permit reproductions for personal use compensated via device levies, against the three-step test. In decisions upholding the levy system under §53 of the German Copyright Act (UrhG), tribunals like the Federal Court of Justice (BGH) have verified step two compliance by confirming that levies prevent conflict with commercial exploitation through fair remuneration, drawing on data showing levy revenues (e.g., €348 million collected in 2014) align with estimated private copying volumes without undermining sales. Challenges arguing overbroad application failed where evidence demonstrated no substitution effect, as levies are calibrated to actual harm via surveys and market studies.^[40] A consistent judicial trend in these rulings favors empirical assessments of market impact for steps two and three, rejecting expansive exceptions lacking quantifiable non-prejudice to rightholders. For instance, CJEU jurisprudence post-Infopaq requires member states to interpret statutory exceptions strictly, ensuring they remain confined to "special cases" via case-specific analysis rather than blanket permissions, thereby safeguarding exploitation rights amid analog-to-digital transitions.^[29]

Extensions to Other Intellectual Property Domains

Comparison with TRIPS Article 30 for Patents

The Berne three-step test, incorporated into Article 13 of the TRIPS Agreement, mandates that copyright exceptions be confined to certain special cases, not conflict with a normal exploitation of the work, and not unreasonably prejudice the legitimate interests of the right holder. By contrast, TRIPS Article 30 authorizes limited exceptions to patent rights only if they do not unreasonably conflict with a normal exploitation of the patent and do not unreasonably prejudice the legitimate interests of the patent owner, while accounting for third-party interests; notably, it omits any requirement to limit exceptions to special cases. This two-pronged structure for patents, lacking the initial confinement criterion, permits greater flexibility for exceptions such as experimental or research use, which are interpreted to support follow-on innovation without commercial substitution.^[41] These divergences stem from fundamental differences in the intellectual property regimes' designs and economic rationales. Patents confer exclusive rights for a fixed term of 20 years from filing, incentivizing high-risk investments in invention and disclosure to enable cumulative technological progress, which justifies exceptions that facilitate verification, regulatory approval, or non-commercial testing without eroding the core monopoly reward. Copyrights, however, extend protection for at least the author's life plus 50 years, safeguarding expressive works with minimal creation barriers and near-zero reproduction costs post-fixed inputs, thus necessitating stricter tests to avert widespread free-riding that could undermine ongoing creative incentives. Scholarly analyses emphasize that the patent provision's balanced consideration of third-party interests—absent in the copyright test—aligns with the disclosure function of patents, promoting knowledge dissemination over the perpetual renewal potential in copyright renewals or derivatives.^[6] Although no direct Berne-style test governs patents internationally, Article 30 functions analogously by prioritizing the patentee's exploitation of research and development returns, as affirmed in WTO disputes like Canada – Patent Protection of Pharmaceutical Products (2000), where experimental use was upheld if non-competitive.^[41] Debates among experts highlight that this looser framework accommodates patents' shorter duration and innovation costs, avoiding the copyright test's rigidity, though some contend it risks underprotecting patentees in emerging fields like biotechnology without explicit "special cases" bounds.^[42]

Limited Relevance to Trademarks and Other Rights

The Berne three-step test, as articulated in Article 9(2) of the Berne Convention for the Protection of Literary and Artistic Works (as revised in 1971), governs exceptions to the exclusive reproduction right in copyright and has been incorporated into Article 13 of the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) to apply to all exclusive rights under copyright. This framework does not extend directly to trademarks, which are regulated separately under TRIPS Article 17; that provision permits "limited exceptions" to trademark rights—such as fair use of descriptive terms—provided they account for the legitimate interests of the trademark owner and third parties, without requiring confinement to special cases or avoidance of conflict with normal exploitation. While Article 17 draws loose inspiration from the three-step test's balancing approach, it employs a more flexible, interest-based criterion rather than the test's cumulative three conditions, reflecting trademarks' distinct emphasis on preventing consumer confusion and protecting source identification.^[43] Exceptions to trademark rights, including those for dilution (prohibited under certain conditions in jurisdictions like the United States via the Trademark Dilution Revision Act of 2006), are evaluated primarily through standards of likelihood of dilution, blurring, or tarnishment, independent of the Berne test's structure. In other intellectual property domains beyond copyrights and patents, such as industrial designs under TRIPS Article 26(2), a parallel but truncated provision allows limited exceptions that do not "unreasonably conflict with the normal exploitation" of the design or "unreasonably prejudice" the owner's legitimate interests, while considering third-party interests; this two-part test operates within designs' field-specific regime and does not incorporate the Berne Convention's full three-step requirements. Judicial and legislative practice demonstrates negligible extension of the Berne three-step test to these areas, maintaining domain-tailored equilibria to avoid imposing copyright-centric constraints on trademarks' or designs' policy objectives, such as functionality doctrines or overlap with competition law.^[8]

Criticisms, Debates, and Emerging Controversies

Claims of Over-Restrictiveness on Exceptions

Critics from user rights advocacy organizations contend that the Berne three-step test unduly constrains copyright exceptions by mandating that they constitute "certain special cases," which necessitates narrowly tailored provisions rather than the broader, case-by-case flexibility afforded by doctrines like U.S. fair use.^[44] This first step, as interpreted in contexts like WTO disputes, requires exceptions to satisfy all three criteria independently, potentially invalidating holistic balancing approaches that weigh multiple factors without rigid thresholds.^[44] The Electronic Frontier Foundation (EFF) has argued that such rigidity in international agreements incorporating the test risks "putting fair use at risk" by entrenching policies that prioritize rightholders over public access, thereby limiting adaptations for digital environments.^[45]^[5] Proponents of expanded exceptions assert that the test hampers innovation, education, and cultural remixing by discouraging uses that do not fit predefined "special" categories, such as transformative digital sharing or non-commercial adaptations.^[5] Knowledge Ecology International (KEI) describes WTO-aligned interpretations as more restrictive than fair use's four-factor analysis, which allows for contextual weighing and supports broader exceptions aligned with promoting science and the arts.^[44] However, these claims of stifled creativity lack robust causal evidence demonstrating net losses in overall content creation or public domain growth attributable to the test's application.^[44] Advocates further maintain that the test's emphasis on avoiding conflict with "normal exploitation" and unreasonable prejudice to rightholders tilts the balance toward corporate interests, constraining the evolution of exceptions needed for technological and societal progress while sidelining incentives for public domain expansion.^[5] EFF positions highlight how this framework in trade pacts like the TPP could narrow access to knowledge and culture, favoring entrenched rightholders over diffuse user benefits.^[45] Such critiques frame the test not as a neutral safeguard but as a barrier to user-centric policies that foster innovation without empirical validation of widespread harm from looser exceptions.^[44]

Defenses Emphasizing Protection of Creators' Interests

Proponents of the Berne three-step test maintain that its criteria—requiring exceptions to be confined to special cases, not conflict with normal exploitation of works, and avoid unreasonable prejudice to authors' legitimate interests—serve as essential safeguards for creators' economic incentives, preventing the erosion of rewards necessary for sustained content production.^[8] By imposing these cumulative constraints, the test counters arguments for unrestricted exceptions, ensuring that any permitted uses do not substitute for licensed markets or diminish authors' ability to derive value from their works.^[3] This framework prioritizes verifiable alignment with the incentive rationale of copyright, where exclusive rights enable creators to recoup upfront investments in expression amid high fixed costs and low marginal reproduction expenses. Historical precedents underscore the risks of inadequate protections, as seen in the pre-Berne Convention era (prior to 1886), when international piracy proliferated due to absent reciprocal rights, severely impacting authors' revenues and discouraging cross-border creation.^[46] Charles Dickens, for instance, earned negligible income from his novels' massive U.S. sales in the 1830s and 1840s, as American publishers reprinted works without payment, prompting his public campaigns for international copyright reform during his 1842 American tour; this exploitation not only reduced his personal earnings but exemplified how unchecked copying stifled incentives for British authors to invest in works vulnerable to foreign markets.^[47] Such episodes illustrate a direct causal connection between weak enforcement against prejudicial uses and diminished creative output, as creators faced systematic free-riding that undermined the financial viability of authorship. Contemporary economic analyses reinforce this linkage, demonstrating that robust copyright frameworks, unhampered by overbroad exceptions, correlate with expanded production in creative sectors.^[48] Copyright-intensive industries contributed approximately 5.3% to global GDP in 2017 and employed over 30 million people worldwide, with evidence indicating that protections against unauthorized exploitation sustain investment in new works by preserving revenue streams from licensing and sales.^[49] Empirical reviews of piracy's analogs—such as uncompensated digital reproductions—show revenue losses exceeding $29 billion annually for the global music industry alone in the early 2010s, correlating with reduced artist investments in recording and promotion until enforcement strengthened.^[50] The three-step test's prejudice criterion thus demands evidence that exceptions impose no such harm, often requiring remuneration mechanisms to maintain creators' legitimate expectations of market-based returns. Critics of expansive exceptions, including creators' organizations, argue that the test rejects ideologically driven dilutions of rights in favor of empirically grounded balances, where normal exploitation encompasses emerging markets like digital licensing, and any deviation must demonstrably not impair authors' control over derivative or transformative uses.^[51] For example, judicial applications in jurisdictions like Germany have upheld the test by conditioning library reproductions on equitable payments, ensuring prejudice is mitigated without ceding core economic interests.^[8] This approach aligns with causal economic reasoning: absent safeguards against unreasonable encroachments, prospective creators anticipate diluted returns, leading to underproduction of public goods like literature and music, as fixed creation costs remain unrecovered amid ubiquitous copying technologies.^[52] Ultimately, the test's rigor fosters a stable ecosystem where incentives drive innovation, verifiable through sustained growth in protected creative outputs rather than unsubstantiated claims of access benefits outweighing losses.

Recent Debates on AI Training Data and Text/Data Mining Exceptions (Post-2020)

The European Union's Directive (EU) 2019/790 on copyright and related rights in the Digital Single Market, transposed into national laws by June 7, 2021, established text and data mining (TDM) exceptions under Articles 3 and 4 to facilitate computational analysis of works. Article 3 mandates an exception for TDM conducted for scientific research purposes by eligible institutions, while Article 4 permits member states to enact an optional exception for other TDM activities, allowing rightholders to opt out via machine-readable means. These provisions explicitly require compliance with the three-step test enshrined in Article 5(5) of Directive 2001/29/EC, which mirrors the Berne Convention's criteria of special cases, no conflict with normal exploitation, and no unreasonable prejudice to legitimate interests. Post-2020 debates have intensified over whether these exceptions adequately support AI model training, particularly for commercial generative systems ingesting vast copyrighted corpora like the Books1 subset of The Pile dataset, which includes scraped books without uniform permission.^[53] Critics contend that Article 4's broad scope fails the first step of the three-step test by not confining exceptions to "certain special cases," as it applies indiscriminately to any TDM developer unless opted out, potentially encompassing mass-scale AI ingestion that resembles general reproduction rather than targeted analysis. This raises causal concerns about undermining emerging data licensing markets, where AI firms have negotiated deals with publishers for training access, suggesting potential conflict with the second step if unlicensed scraping supplants remunerated exploitation. Empirical assessments of harm remain sparse, but evidence of AI models regurgitating verbatim excerpts from training texts has fueled arguments that such uses unreasonably prejudice authors' interests under the third step, absent robust opt-out enforcement or transparency mandates. Proponents counter that transformative training—where inputs yield non-expressive outputs—avoids direct substitution, though courts have emphasized case-specific evaluation beyond mere TDM labeling.^[53]^[54] In 2023–2024, European lawsuits tested these tensions, exemplified by the Hamburg Regional Court's September 27, 2024, ruling in a case against LAION e.V. for scraping public images into AI datasets. The court held that non-commercial dataset creation for research-oriented AI training qualified under national implementations of the Article 3 exception, deeming it a special case compliant with the three-step test despite scale, as no evidence showed conflict with exploitation or prejudice—though it distinguished this from downstream commercial model deployment. Similar proceedings, including opt-out disputes, highlight unresolved issues like the practicality of reservations for pre-existing datasets and whether commercial TDM inherently breaches step two amid growing licensing precedents. These cases underscore a judicial reluctance to deem mass ingestion per se violative without proven harm, yet reveal fractures in harmonization across member states.^[55]^[56] International forums have amplified calls for evidence-based calibration, with WIPO's Standing Committee on Copyright and Related Rights (SCCR/44, October 2023) discussing whether expansive TDM for AI satisfies the Berne test's "special cases" threshold, advocating narrow definitions tied to verifiable research benefits over indiscriminate commercial application. Organizations like CISAC have warned that unchecked TDM risks eroding the three-step balance, urging empirical studies on market displacement—such as comparative sales data pre- and post-AI proliferation—before broadening exceptions globally. WTO panels' prior interpretations of the test, emphasizing non-hypothetical conflicts, inform these debates, yet consensus eludes due to divergent national implementations and limited cross-jurisdictional data on causal impacts like reduced derivative licensing revenues. Ongoing WIPO AI-IP conversations post-2020 stress transparency in datasets to enable prejudice assessments, potentially bridging gaps without diluting core protections.^[57]^[58]

References

[1]
https://www.wipo.int/treaties/en/text.jsp?file_id=283698
[2]
The Role of the three-step test in the adaptation of copyright law to ...
The three-step test appeared for the first time in 1967 at the Stockholm Conference, whose objective was to revise the Berne Convention.<|separator|>
[3]
[DOC] Limitations and Exceptions under the "Three-Step Test&quot ...
The Stockholm Conference provision on exceptions to the reproduction right, the three-step test, eventually passed into the 1971 Paris Act of the Berne ...
[4]
[PDF] The Three-Step-Test Revisited: How to Use the Testâ•Žs Flexibility ...
THE THREE-STEP TEST IN THE BERNE CONVENTION. The first three-step test in international copyright law emerged as. Article 9(2) of the Berne Convention. It ...
[5]
[PDF] Three-‐Step Test
For example, whereas the. Berne Convention three-‐step test only applies to exceptions and limitations to the right of reproduction, the three-‐step test ...
[6]
A Scientific Analysis of the Three-Step Test: Through the Lenses of ...
Jul 5, 2022 · In respect of free use for reproduction, the Berne Convention contains a general rule rather than an explicit limitation: art 9(2) of the Berne ...
[7]
https://www.wipo.int/export/sites/www/copyright/en/activities/pdf/stockholm_conference_1967.pdf
[8]
None
Summary of each segment:
[9]
The Berne Convention revisions for limitations and exceptions to ...
This research note reports on the evolution of each copyright exception found in the Berne Convention for the Protection of Literary and Artistic Works.
[10]
The 1967 Stockholm Revision of the Berne Convention Protocol ...
Aug 12, 2012 · A Protocol was approved that provided for much greater access to copyrighted works than the earlier 1948 revision of the Berne Convention.
[11]
[PDF] Guide to the Berne Convention for the Protection of Literary ... - WIPO
The Berne Convention for the Protection of Literary and Artistic Works was concluded on September 9, 1886. It is the oldest of the international copyright ...
[12]
Berne Convention for the Protection of Literary and Artistic Works
### Summary of Article 9(2) Three-Step Test in the 1971 Paris Act
[13]
WIPO Lex
Summary of each segment:
[14]
WIPO Lex
Contracting Parties/Signatories Berne Convention (Total Members: 182). Member, Signature, Instrument, In Force, Details. Afghanistan.
[15]
BERNE CONVENTION, AS REVISED - Article 9 - Law.Cornell.Edu
(2) It shall be a matter for legislation in the countries of the Union to permit the reproduction of such works in certain special cases, provided that such ...
[16]
intellectual property (TRIPS) - agreement text - standards - WTO
TRIPS standards include copyright (complying with Berne Convention), computer programs, 50-year protection term, and trademark protection for signs ...
[17]
[PDF] guide to the copyright and related rights treaties administered by wipo
Paragraph (2): “three-step test” for exceptions and limitations. BC-9.11. Paragraph (2) corresponds to the requirement stated in the very early stages of the ...
[18]
[PDF] A scientific analysis of the three-step test - Edith Cowan University
Jun 21, 2022 · The three-step test emerged in international copyright law; it is embodied not only in the Berne Convention [art 9(2)] but also in the TRIPs ...
[19]
[PDF] A Comparative Analysis of the Three-Step Tests in International ...
... detailed comparison of each element of each step, across all ... force4 that contain a three-step test: the Berne Convention (“Berne”);5 the WIPO Copyright.
[20]
[PDF] How to Use the Test's Flexibility in National Copyright Law
2/RES/2001/21). 69 According to Martin Senftleben, conflict with normal exploitation should occur only if “the author is deprived from a current or ...
[21]
[PPT] Applying the three step test in the digital environment - WIPO
The reproduction right, as set out in Article 9 of the Berne Convention, and the exceptions permitted thereunder, fully apply in the digital environment, in ...
[22]
[PDF] The Reverse Three-Step Test - Scholarship@Vanderbilt Law
The paper will then suggest that a new international copyright norm could be created based on the Berne Convention's three-step test, in harmony with the U.S..
[23]
intellectual property - overview of TRIPS Agreement - WTO
Article 13 requires Members to confine limitations or exceptions to exclusive rights to certain special cases which do not conflict with a normal exploitation ...
[24]
https://www.cambridge.org/core/books/general-exception-clauses-of-the-trips-agreement/determining-the-normative-meaning-of-art-13-of-the-TRIPS-agreement-and-art-92-of-the-berne-convention/9A55E470E40C7A7CC3160BFAE1EA1A7A
[25]
Summary of the WIPO Copyright Treaty (WCT) (1996)
As to limitations and exceptions, Article 10 of the WCT incorporates the so-called "threestep" test to determine limitations and exceptions, as provided for ...
[26]
Summary of the WIPO Performances and Phonograms Treaty ...
As to limitations and exceptions, Article 16 of the WPPT incorporates the so-called "threestep" test to determine limitations and exceptions, as provided ...
[27]
WIPO Lex
WIPO-Administered Treaties. Export to Excel. Contracting Parties/Signatories WIPO Copyright Treaty (Total Members: 118) ...
[28]
https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32001L0029
[29]
https://academic.oup.com/grurint/article/72/7/631/7071647
[30]
DS160 United States — Section 110(5) of US Copyright Act
The dispute centred on the compatibility of two exemptions provided for in Section 110(5) of the US Copyright Act with Article 13 of the TRIPS Agreement.Missing: three- step test
[31]
[PDF] 1234da.pdf - World Trade Organization
Jun 15, 2000 · 6.16. Thus we conclude that it is for the European Communities to present a prima facie case that. Section 110(5)(A) and (B) of the US Copyright ...
[32]
[PDF] Toward Supranational Copyright Law? The WTO Panel Decision ...
This article will analyze the Panel's interpretation of the test=s three steps, and their application to the U.S.-law exemption. The article will also compare ...
[33]
[PDF] wto panel report on section 110(5) of the
The EC alleged that sections 110(5)(A) and (B) of the. United States Copyright Act, as amended by the FMLA, violated Article. 9.1 of the TRIPS agreement and ...
[34]
[PDF] 18-1 CHAPTER 18 INTELLECTUAL PROPERTY Section A - USTR
Dec 20, 1996 · Berne Convention means the Berne Convention for the Protection of Literary ... 3 (Protection of Undisclosed Test or Other Data), if a product is ...Missing: CPTPP | Show results with:CPTPP
[35]
[PDF] US Proposal for an Intellectual Property Chapter in the Trans
three-step test and its inclusion in the TPP. First, unlike in the Berne. Convention, but similar to Article 13 of TRIPS, the clause in the TPP proposal ...
[36]
[DOC] SCCR/9/7: WIPO Study on Limitations and Exceptions of Copyright ...
“The reproduction right, as set out in Article 9 of the Berne Convention, and the exceptions permitted thereunder, apply fully in the digital environment ...
[37]
Is the U.S. Fair Use Doctrine Compatible with Berne and TRIPS ...
Aug 22, 2018 · The paper argues the U.S. fair use doctrine is compatible with Berne and TRIPS, satisfying the three-step test, and has been consistent since ...
[38]
Fair use complies with the three-step test | ALRC
Dec 2, 2013 · 4.140 The three-step test was first incorporated into international copyright law during the 1967 Stockholm revision of the Berne Convention.
[39]
https://curia.europa.eu/juris/document/document.jsf?docid=72620&doclang=EN
[40]
https://wipo.int/edocs/pubdocs/en/wipo_pub_1037_2016.pdf
[41]
[PDF] WTO ANALYTICAL INDEX TRIPS Agreement – Article 30 (DS reports)
Article 30 allows limited exceptions to patent rights if they don't unreasonably conflict with normal exploitation, prejudice patent owner's interests, and are ...
[42]
(PDF) A Comparative Analysis of the Three-Step Tests in ...
Aug 6, 2025 · ... conflict with normal exploitation). Provision. Link to Step 1. Condition. Protected Action. Object. Link to. Step 3. Berne Art. 9(2). provided ...
[43]
2 - A Look at the Trademark Provisions in the TRIPS Agreement
Its wording is inspired from the three-step test found in Article 13 of TRIPS concerning exceptions and limitations to copyright rights, which itself comes from ...<|separator|>
[44]
United States Four Fair Use Factors and the WTO Three-Step Test
Nov 21, 2012 · If one factor is not satisfied, the inquiry ends and the limitation or exception will be found in non-compliance with the three-step test.
[45]
New Leaked TPP Text Puts Fair Use at Risk
Aug 3, 2012 · It is critical to ensure that any language introducing the three-step test ensure that the test is not narrowly construed and also ensure ...
[46]
Charles Dickens And Copyright Law: Five Things You Should Know
Jan 17, 2017 · At this time, American pirating of English (and other foreign) authors was common, perfectly legal, and considered to be consistent with ...
[47]
Bootlegging Dickens: Author Looks At 'Bookaneers' - NPR
even the biggest, like Charles Dickens.
[48]
The Economic Performance of Copyright-Based Industries - WIPO
Industries that make use of copyright and related rights protection are often referred to as the copyright-based or the creative industries.
[49]
[PDF] AN EMPIRICAL ANALYSIS OF THE ECONOMICS OF COPYRIGHT
The principal objective is to attempt to provide guidance to economists interested in undertaking empirical research on the economics of copyright, particularly ...
[50]
[PDF] The Resilience of Creativity - U.S. Copyright Office
Notably, the production of registered creative works recovered well before impacted employees went back to work within copyright-reliant industries.
[51]
[PDF] PERSPECTIVES UNDER THE BERNE THREE-STEP TEST AND ...
Sep 26, 2024 · This is the reward the copyright system gives authors: the power to solely exploit their work as they see fit. For example, creators may license ...
[52]
[PDF] Identifying the Economic Implications of Artificial Intelligence for ...
Generally, offering copyright protection to AI-generated works will further incentivize their production, exacerbating the effects, both positive and negative, ...
[53]
[PDF] Evasion of Copyright Protection via the EU's Text and Data Mining ...
Though Berne's version of the three steps focuses on individual countries' applications of the test, the TRIPS Agreement focuses on members' obligation to ...Missing: variations | Show results with:variations<|separator|>
[54]
These 183,000 Books Are Fueling the Biggest Fight in ... - The Atlantic
Sep 25, 2023 · This summer, I acquired a data set of more than 191,000 books that were used without permission to train generative-AI systems by Meta, ...
[55]
To Scrape or Not to Scrape? First Court Decision on the EU ...
Oct 4, 2024 · [1] The three-step test is laid down in Art. 5(5) of the EU InfoSoc Directive 2001/29/EC that also applies to the new copyright exceptions ...
[56]
German Court Says Non-Commercial AI Training Data Meets ...
Oct 10, 2024 · The case focuses on whether copying an image to create an AI training dataset constitutes copyright infringement, not on AI model training or ...
[57]
[PDF] Standing Committee on Copyright and Related Rights - WIPO
Oct 17, 2023 · According to the three-step test, L&E for research purposes must be carefully defined to avoid exempting uses in a manner that conflicts with ...
[58]
CISAC Statement for the 9th Session of WIPO's Conversations on IP ...
Mar 14, 2024 · Such application of TDM exceptions may undermine the principles of the three-step test and disrupt the delicate balance inherent in copyright ...Missing: WTO Berne