News

Patent, liability and know-how disclosure: What steps should you take to stay protected?

In today’s innovation-driven economy, where collaborative research and technical partnerships are increasingly common, the exchange of proprietary know-how has become both essential and legally sensitive. If not properly managed, the disclosure of confidential technical information can compromise patentability, lead to the forfeiture of intellectual property rights, or give rise to liability for the disclosing or receiving party. This article explores the legal risks associated with know-how disclosure in the patent context illustrated by recent case law and sets out practical safeguards to ensure such exchanges are secure and compliant.

Legal risks arising from know-how disclosure

When confidential disclosure leads to patent litigation

The disclosure of technical know-how, even in a collaborative setting, may give rise to complex legal claims if improperly managed. Risks include:

  • Ownership claims by the original holder of the disclosed information;
  • Breach of contract claims, especially in the presence of a non-disclosure agreement (NDA);
  • Irretrievable loss of confidentiality, should the disclosed elements be publicly revealed through a patent application.

Failure to anticipate these risks can not only nullify competitive advantage but also result in judicial reassignment of the patent rights and financial liability.

Case study: Air Liquide (Paris Judicial Court, January 31, 2025)

In 2020, Futura Mechanical Design Project (FMDP) and its parent company F2M SAS were retained by Air Liquide to conduct a feasibility study on a liquid hydrogen pump. An NDA was signed, explicitly forbidding the recipient from using disclosed technical information to file any intellectual property rights.

Despite this, an affiliated company within the Air Liquide group filed a French patent and subsequently a PCT international application, the contents of which replicated confidential material from the commissioned study. FMDP and F2M subsequently brought a claim for ownership of the patent, arguing that their proprietary know-how and the outcomes of their technical study had been unlawfully appropriated and used without authorization.

The Paris Judicial Court found that:

  • FMDP held the technical solution prior to the collaboration;
  • The patent filing constituted a breach of the NDA;
  • The General Terms and Conditions (GTCs) cited by Air Liquide, which purportedly transferred IP ownership, were neither agreed upon nor enforceable.

The Court:

  • Ordered the transfer of the patent and its international extensions to FMDP and F2M SAS;
  • Awarded €30,000 in damages;
  • Rejected Air Liquide’s counterclaims as unfounded and untimely.

This case underscores the evidentiary value of prior ownership documentation, the contractual weight of confidentiality obligations, and the inadmissibility of relying on unaccepted boilerplate GTCs in matters of intellectual property.

Contractual confidentiality obligations and legal consequences

A well-drafted NDA is a critical safeguard in collaborative innovation. Standard confidentiality provisions generally prohibit:

  • The use of disclosed information beyond the defined scope;
  • The filing of any intellectual property rights (patent, utility model, etc.) derived from such disclosures.

Failure to comply may result in:

  • Injunctions and court-ordered ownership reassignment of the IP in question;
  • Monetary damages for loss and reputational harm;
  • Potential exposure of trade secrets through patent publication, which is irreversible.

Best practices when disclosing know-how in patent-related contexts

  1. Draft robust and purpose-specific non-disclosure agreements NDAs

Generic templates are often insufficient. Tailor each NDA to the context, ensuring:

  • A clear definition of what constitutes confidential information;
  • Explicit terms regarding the prohibition (or permitted use) of such information in patent filings;
  • Provisions reserving all pre-existing intellectual property rights of the disclosing party.
  1. Secure proof of prior ownership of the invention

Maintain a systematic internal record of technical developments through:

  • Invention Disclosure Forms (IDFs), signed and timestamped;
  • Digital vaults or sealed e-Soleau envelopes for sensitive developments;
  • Logs of internal emails and R&D notes to evidence early-stage ownership.

These records are crucial in proving entitlement and defeating wrongful ownership claims.

  1. Clarify intellectual property clauses in all agreements

Beyond NDAs, ensure contracts such as R&D orders, joint development agreements, or licensing deals clearly state:

  • Who owns the results and deliverables;
  • The extent and limitations of assignments or licensing rights.
  1. Explicitly reject conflicting general terms and conditions

Never rely on silence. In the Air Liquide case, the GTCs invoked by one party were ultimately deemed unenforceable due to:

  • Lack of acceptance by the counterparty;
  • Inconsistent communications showing clear disagreement.

To avoid ambiguity:

  • Always confirm acceptance or rejection of GTCs in writing;
  • Include IP-specific carve-outs in contracts or appendices.
  1. Implement a controlled internal disclosure protocol

Companies should adopt a standardized protocol prior to any external exchange:

  • Legal review of disclosure scope;
  • Document classification and labelling (e.g., “Confidential – Do Not Distribute”);
  • Internal tracking of disclosures: to whom, when, why, and under what conditions.

Conclusion: minimizing patent liability through preventive action

Patent-related liability arising from improper use or disclosure of know-how can be devastating legally, financially, and reputationally. Whether in open innovation or bilateral technical cooperation, prevention through contractual rigor and internal diligence remains the best defense.

Dreyfus Law Firm works with clients in the food sector, providing specialist advice on intellectual property and regulatory issues to ensure compliance with national and European laws.

We collaborate with a global network of intellectual property attorneys.

Join us on social media !

LinkedIn  

Instagram 

FAQ

1. Can an invention disclosed under an NDA be patented?

No, unless the NDA explicitly authorizes it and the information does not qualify as protected know-how.

2. What if both parties contribute to the same invention?

A co-inventorship agreement or a joint filing strategy must be agreed upon before any patent application is submitted.

3. Are GTCs alone sufficient to secure IP rights?

No. Without clear and formal acceptance, GTCs have no binding force over intellectual property rights.

Read More

UNITED KINGDOM: Width and size of a specification can be indicative of bad faith

The UK Supreme Court’s decision in SkyKick UK Ltd and another v. Sky Ltd and others has significantly impacted trademark law, particularly concerning the breadth of specifications and the concept of bad faith. This ruling underscores the importance of aligning trademark specifications with genuine commercial intentions.

Background of the SkyKick v Sky Case

In 2016, Sky Ltd, a prominent broadcaster and telecommunications company, initiated legal proceedings against SkyKick, a US-based cloud management software provider, alleging trade mark infringement. Sky’s claims were based on its extensive trade mark registrations covering a wide array of goods and services. SkyKick countered by challenging the validity of Sky’s trademarks, asserting that they were registered in bad faith due to their overly broad specifications without genuine intent to use the marks across all listed categories.

Legal framework and Supreme Court’s findings

Understanding bad faith in trademark law

Under Section 3(6) of the UK Trade Marks Act 1994, a trademark shall not be registered if the application is made in bad faith. The concept of bad faith involves a lack of genuine intention to use the trade mark for the goods and services specified at the time of application.

Supreme Court’s analysis

The Supreme Court held that:

  • Overly broad specifications: Filing for a wide range of goods and services without a genuine intention to use the trade mark for all of them can indicate bad faith.
  • Assessment of intent: The applicant’s intention at the time of filing is crucial. A lack of intention to use the mark for certain goods or services, especially when the specification is excessively broad, supports a finding of bad faith.
  • Partial invalidity: If bad faith is established for certain goods or services, the trade mark can be partially invalidated for those specific categories.

This decision emphasizes that trade mark applications must reflect a genuine commercial strategy and not serve as a means to unjustly monopolize market segments.

Implications for trademark applicants

Strategic considerations

Trademark applicants should:

  • Align specifications with business activities: Ensure that the goods and services listed in the application correspond to current or planned business operations.
  • Avoid overly broad terms: Refrain from using vague or broad categories without a clear intention to use the trade mark across all specified areas.
  • Maintain documentation: Keep records demonstrating the intention to use the trade mark for each specified good or service at the time of application.

Risk management

Companies must be aware that:

  • Enforcement actions may backfire: Initiating infringement proceedings based on broad specifications can lead to counterclaims of bad faith, potentially resulting in partial invalidation of the trade mark.
  • Portfolio audits are essential: Regularly reviewing trade mark portfolios to ensure that all registrations are defensible and align with genuine business intentions is crucial.

Conclusion

The Supreme Court’s ruling in the SkyKick case serves as a pivotal reminder of the importance of integrity and genuine intent in trade mark applications. Applicants must ensure that their trade mark specifications are precise and reflect actual or planned use to withstand legal scrutiny and maintain robust protection.

 

At Dreyfus Law Firm, we specialize in intellectual property law, offering expert guidance on trade mark registration, portfolio management, and enforcement strategies. Our team is dedicated to ensuring that your intellectual property rights are secured and maintained with the highest level of professionalism and integrity.

Dreyfus Law Firm is in partnership with a global network of attorneys specializing in Intellectual Property.

Join us on social media !

LinkedIn  

Instagram

FAQ

What constitutes bad faith in trade mark applications?

Bad faith occurs when an applicant files a trade mark application without a genuine intention to use the mark for the specified goods or services, often to prevent others from entering the market or to gain an unfair advantage.

Can a trade mark be partially invalidated for bad faith?

Yes, if bad faith is established for certain goods or services within a trade mark specification, the registration can be partially invalidated for those specific categories.

How can I ensure my trade mark application is not considered in bad faith?

Align your trade mark specifications with your current or planned business activities, avoid overly broad terms, and maintain documentation demonstrating your intention to use the mark for each specified good or service.

Read More

NIS360: A collaborative approach to strengthening cyber risk management

In an era marked by increasingly sophisticated and cross-sectoral cyberattacks, organizations must move beyond fragmented responses. A structured, anticipatory, and collaborative approach to cyber-risk governance is now imperative.

The NIS360 framework, advocated by Ioanna Antcheva, emerges as a strategic solution—an approach grounded in information sharing, strong governance, and continuous improvement. At its core lies a fundamental belief: digital security is a collective responsibility.

NIS360: an integrated cybersecurity framework

Designed for both public and private organizations, NIS360 offers a holistic methodology to govern cyber risks through three essential dynamics: identification, anticipation, and response.

Foundational pillars of NIS360

  • Risk identification and control

The process begins with mapping internal vulnerabilities and external threats while evaluating their potential business impact. This diagnostic phase is critical to designing an effective and resilient cybersecurity architecture.

  • Incident response planning

Organizations must prepare for the inevitable. This includes developing real-time response protocols, assigning clear operational roles, and conducting simulations to test crisis readiness.

  • Strategic intelligence sharing

A core principle of NIS360 is to foster structured collaboration between public institutions, private entities, and regulators. Shared cyber intelligence accelerates threat detection and enhances collective resilience.

  • Compliance and governance

Adherence to regulatory frameworks such as the NIS2 Directive, GDPR, and national cybersecurity agency guidelines (e.g., ANSSI in France) must be embedded into corporate governance. Executive leadership must be directly involved in cyber oversight.

Operationalizing the NIS360 framework

Engaging all stakeholders

Cybersecurity cannot be siloed within IT departments. The NIS360 model calls for the active engagement of all business units executive management, legal, human resources, procurement, and external partners. This transversal alignment strengthens coherence and accountability.

Continuous monitoring and adaptive response

A cybersecurity framework must evolve alongside the threat landscape. Long-term effectiveness relies on:

  • Timely updates to detection and prevention tools
  • Routine audits and performance assessments
  • Agility in updating policies, controls, and procedures

NIS360 promotes a “cybersecurity lifecycle” approach, incorporating continuous legal and technological monitoring to adapt to emerging risks.

Legal and regulatory considerations

Implementing the NIS360 framework requires close attention to legal risk management. Key compliance sources include:

  • The NIS2 Directive, which expands obligations for essential and important entities across critical sectors
  • The GDPR, particularly concerning breach notification and data protection principles
  • National-level recommendations (e.g., CNIL, ANSSI) that detail preventive measures and incident response protocols

Failure to comply may expose an organization to administrative fines, reputational harm, and even civil or criminal liability.

Conclusion and strategic outlook

The NIS360 framework establishes a new European benchmark in cyber-risk management. It encourages organizations to embrace a proactive, integrated, and leadership-driven approach.

Anticipation, information sharing, and compliance are the cornerstones of this model. Organizations that embed NIS360 not only strengthen their cyber-resilience but also bolster their market credibility with stakeholders, regulators, and investors.

Dreyfus Law Firm works with clients in the food sector, providing specialist advice on intellectual property and regulatory issues to ensure compliance with national and European laws.

We collaborate with a global network of intellectual property attorneys.

Join us on social media!

LinkedIn  

Instagram 

FAQ

What is the NIS360 framework?

NIS360 is a structured framework for cyber-risk governance, based on strategic intelligence sharing, organizational resilience, and legal compliance.

What are the main legal obligations in cybersecurity?

Under the NIS2 Directive and GDPR, entities must report major incidents, protect personal data, and implement organizational and technical safeguards.

How can an organization establish effective cybersecurity governance?

Designate a cybersecurity officer, integrate risk management into core business operations, and monitor KPIs to ensure ongoing performance.

Read More

EU Design Law Reform: What It Means for Rights Holders — Insights from the Dreyfus law firm in Paris

On May 1st, 2025, the European Union officially launched the first phase of its long-awaited design law reform, with significant implications for businesses across the EU. While these changes modernize design protection — particularly in digital and fast-moving industries — they also introduce steep renewal fees, prompting companies to rethink their long-term strategy.

At Dreyfus, we’ve been supporting businesses for over 20 years in securing and optimizing their intellectual property. Our founding partner Nathalie Dreyfus was interviewed by MLex to share her insights on this reform and how companies can prepare.

Sharp Rise in Renewal Fees

The most immediate change is the increase in renewal fees. For example, renewing a single design now costs €150 for the first renewal (up from €90), and can rise to €700 for the fourth renewal.

“The significant increase in renewal costs poses a challenge, particularly for companies managing extensive portfolios,”
Nathalie Dreyfus, MLex interview

We advise clients to conduct a portfolio audit to determine which designs should be maintained long term and which could be streamlined or reorganized.

Download the original article in PDF: Click here to view

A Chance to Optimize Filing Strategy

While renewals are more expensive, the reform offers several improvements:

  • Lower initial application fees: €250 flat fee now includes both registration and publication.
  • Up to 50 designs per application allowed, with no grouping by product type.
  • Explicit protection for digital and 3D designs, including app interfaces and 3D print files.

At Dreyfus, we help clients adapt through proactive audits, portfolio optimization, and cost-efficient international filings, including via the Hague System.

Easier Access to Invalidation for SMEs?

A major feature — set to launch in 2027 — is the administrative invalidation procedure, designed to offer a faster, more affordable alternative to court-based challenges.

“To ensure fairness and legal certainty, administrative invalidation should be mandatory across all EU member states,”
Nathalie Dreyfus, MLex

Our team supports a harmonized approach to ensure SMEs can access efficient legal remedies regardless of jurisdiction.

Why Work with Dreyfus in Paris?

  • Tailored strategies – We analyze your business model, market targets, and product lifecycle to build the right protection plan.
  • International scope – We secure your designs in Europe and beyond.
  • Proactive legal monitoring – We continuously monitor legislative changes to keep your IP strategy ahead of the curve.

FAQ – EU Design Reform 2025

What are the new renewal fees?
€150 for the first renewal, up to €700 for the fourth. This reflects the EU’s goal of encouraging selective long-term protection.

Can I still submit multiple designs in one application?
Yes — up to 50 designs can be filed in a single application, without needing to group them by type.

Are digital and virtual designs now protected?
Yes — a key innovation of the reform. Designs like app interfaces, animated elements, and 3D files now enjoy explicit legal protection.

What’s changing in 2026 and 2027?
Administrative invalidation will roll out in July 2026. The “repair clause” — impacting spare parts and competition — takes effect in December 2027.

How can I decide which designs to renew or abandon?
We provide customized portfolio audits to assess the commercial value of each design and guide your strategic decisions.


Want to protect your design portfolio or plan for the new EU rules?
Contact our team to develop a secure, forward-looking strategy.

 

Read More

Fashion law in France: A strategic legal framework for the luxury industry

Fashion law in France forms a unique legal architecture designed to protect and promote the excellence of French craftsmanship. In a global market driven by creativity and innovation, understanding the legal rules governing manufacturing, distribution, and intellectual property is essential for fashion and luxury brands. This legal corpus enables France to maintain a leading role in both innovation and protection.

A robust legal framework for manufacturing and supply

 Applicable regulations

In France, the manufacturing of fashion products is primarily governed by general contract law, as set forth in Articles 1101 et seq. of the French Civil Code. However, this framework is complemented by sector-specific rules, particularly regarding:

  • Use of specific materials: Legislation regulates the use of fur, exotic leather, and diamonds in accordance with international conventions such as CITES.
  • Product origin: The “Made in France” label is subject to strict criteria. To lawfully use this indication, the product must be mainly manufactured and assembled in France. The DGCCRF ensures compliance with these rules, and French Customs carries out additional checks.

In addition, Haute Couture, the emblem of French luxury, is governed by legal standards. The term is reserved for fashion houses accredited by the Chambre Syndicale de la Haute Couture, which must, among other things, operate at least one workshop in Paris employing a minimum of 20 skilled seamstresses, produce made-to-measure garments, and present a collection of at least 25 original models each season.

Contracts used throughout the value chain

Relationships between brands, manufacturers, subcontractors, and suppliers are formalised through detailed commercial contracts. The most common include :

  • Manufacturing or assembly agreements: Define production methods, quality standards, timelines, and liabilities.
  • Subcontracting agreements: Specify the subcontractor’s obligations, particularly regarding confidentiality and intellectual property compliance.
  • Supply or purchase contracts: Include commitments on volumes, specifications, and conformity.

Each contract should contain clear provisions regarding deadlines, responsibilities, dispute resolution mechanisms, penalties for defects, and ownership of intellectual property rights over the creations.

Distribution models and commercial agency agreements

Selective, exclusive or open: Common structures

Distribution systems in France are governed by both French Commercial Law and EU regulations, particularly Regulation (EU) 2022/720 on vertical agreements. This regulation provides certain exemptions from the general prohibition of anti-competitive agreements, particularly for selective distribution, which is widely used in the luxury sector.

The most common models include:

  • Selective distribution: The supplier appoints resellers based on objective quality criteria to maintain the brand’s image.
  • Exclusive distribution: A single distributor is designated for a specific geographic area and benefits from exclusive rights.
  • Open (non-exclusive) distribution: Each party retains significant contractual freedom.

Specific rules governing commercial agents

The status of commercial agents is strictly governed by Articles L134-1 et seq. of the French Commercial Code. A commercial agent is an independent intermediary who negotiates and, in some cases, concludes contracts on behalf of the principal.

This status includes protective legal measures, such as:

  • Mandatory notice period before termination
  • Compensatory indemnity in the event of termination (except in case of gross misconduct)
  • Registration requirement with the relevant commercial court

Imports, exports and sensitive materials

Fashion goods are subject to European Union customs rules. Customs duties generally range around 12% for garments and accessories imported from outside the EU.

France also implements key international frameworks, such as:

  • CITES, which regulates trade in materials derived from protected species (e.g. alligator or python leather)
  • The Kimberley Process, which governs the ethical trade of diamonds

Additionally, EU economic sanctions restrict certain exports, notably of luxury goods to Russia and Belarus.

Intellectual property: At the core of brand value

Legal protection of creations is vital in the fashion sector. Several tools are available :

  • Trademark law : Protects the brand name, logo, slogans, and distinctive visual elements.
  • Design rights : Registered with the INPI or EUIPO, protect the appearance and aesthetic features of products.
  • Copyright : Automatically protects original designs (cuts, patterns), without requiring registration.

A well-designed intellectual property strategy involves a combination of trademark registration, design filings, and active monitoring across distribution channels.

Conclusion

Fashion law in France offers companies a rigorous and protective legal foundation, both to secure their production chains and to safeguard the commercial and artistic value of their creations.

To remain competitive and legitimate, brands must integrate appropriate legal tools at each stage of their development, from product design to distribution, ensuring full legal protection of their business model and brand identity.

Dreyfus Law Firm offers tailored brand protection strategies rooted in IP law, with a proactive and international scope.

Dreyfus Law Firm is in partnership with a global network of Intellectual Property law specialists.

Join us on social media !

LinkedIn  

Instagram

FAQ

What are the legal requirements for manufacturing and using the "Made in France" label ?

Manufacturing is governed by general contract law (French Civil Code) and specific regulations. To bear the "Made in France" label, a product must be primarily manufactured and assembled in France. Compliance is monitored by the DGCCRF and French customs authorities.

What distribution models are permitted for luxury fashion brands ?

Three main structures are recognised : selective distribution (distributors chosen based on objective criteria), exclusive distribution (a single distributor appointed for a defined territory), and non-exclusive (free) distribution. These models must comply with both French law and EU Regulation 2022/720 on vertical agreements.

What legal protections are available for fashion creations in France ?

Fashion creations are protected through trademarks, designs, and copyright. Trademarks and designs require registration (INPI, EUIPO), while copyright applies automatically to original creations. A comprehensive IP strategy combining these tools is essential to prevent and combat counterfeiting.

Read More

Navigating UK trademark use derived from EU trademark registrations post-Brexit

The UK’s departure from the European Union has significantly altered the landscape of trademark protection and enforcement. Businesses must now navigate a bifurcated system where EU trademarks (EUTMs) no longer extend protection to the UK, and vice versa. This article aims to provide a comprehensive overview of the current state of UK trademark use derived from EU trademark registrations, highlighting key considerations for rights holders operating across these jurisdictions.

Understanding the post-Brexit trademark landscape

 Separation of jurisdictions

As of January 1, 2021, EUTMs no longer confer protection within the UK. To mitigate the impact on rights holders, the UK Intellectual Property Office (UKIPO) automatically created comparable UK trademarks for all existing EUTMs. These comparable rights retain the original filing and priority dates of the corresponding EUTMs but are now subject to UK law and jurisdiction.

Scope of protection

The separation means that EUTMs cover only the remaining EU member states, while UK trademarks (including comparable UK rights) cover only the UK. Therefore, businesses seeking comprehensive protection across both regions must maintain separate registrations.

Use requirements for EUTMs and UK trademarks

 Genuine use pre-Brexit

For both EUTMs and comparable UK trademarks, use of the mark in the UK prior to January 1, 2021, is considered valid for demonstrating genuine use within the respective territories. This means that pre-Brexit use in the UK can support the validity of an EUTM and vice versa.

Genuine use post-Brexit

Post-Brexit, the requirements diverge:

  • EUTMs: Use of the mark must occur within the EU member states to maintain the registration. Use solely in the UK after January 1, 2021, does not suffice.
  • UK trademarks: Use of the mark must occur within the UK. Use solely within the EU post-Brexit is insufficient to demonstrate genuine use in the UK.

Implications for trademark enforcement and strategy

Risk of revocation

Failure to demonstrate genuine use within the appropriate jurisdiction can lead to revocation of the trademark. For instance, a comparable UK trademark not used in the UK within a continuous five-year period post-Brexit is vulnerable to cancellation.

Strategic considerations

Businesses must reassess their trademark portfolios to ensure that they have adequate protection in both the UK and EU. This may involve:

  • Filing new applications in the UK or EU to cover gaps in protection.
  • Monitoring use of marks to ensure compliance with genuine use requirements.
  • Adjusting enforcement strategies to account for the separate jurisdictions.

Best practices for rights holders

To navigate the post-Brexit trademark environment effectively, rights holders should:

  • Audit existing portfolios: Review current registrations to identify any vulnerabilities due to lack of use in the appropriate jurisdiction.
  • Maintain separate registrations: Ensure that trademarks are registered separately in the UK and EU to maintain protection across both regions.
  • Monitor use: Keep detailed records of where and how trademarks are used to support claims of genuine use.
  • Seek professional advice: Consult with trademark attorneys to develop strategies tailored to the business’s specific needs and markets.

Conclusion

The divergence of UK and EU trademark systems post-Brexit necessitates a proactive approach to trademark management. Rights holders must ensure that they have appropriate registrations and can demonstrate genuine use within each jurisdiction to maintain and enforce their trademark rights effectively.

Dreyfus Law Firm specializes in intellectual property law and offers comprehensive services to assist businesses in navigating the complexities of trademark protection in the UK and EU. We are partnered with a global network of attorneys specializing in intellectual property to provide our clients with seamless support across jurisdictions.

Dreyfus Law Firm is partnered with a global network of attorneys specializing in intellectual property law.

Join us on social media !

LinkedIn  

Instagram

FAQ

Does my EUTM still protect my trademark in the UK?

No, as of January 1, 2021, EUTMs no longer provide protection in the UK. However, the UKIPO has created comparable UK trademarks for existing EUTMs to maintain protection within the UK.

Can I rely on use of my trademark in the UK to maintain my EUTM?

Only if the use occurred before January 1, 2021. Post-Brexit use in the UK does not count towards maintaining an EUTM.

What constitutes genuine use of a trademark?

Genuine use involves actual use of the trademark in the market for the goods or services for which it is registered. This includes sales, advertising, and other commercial activities demonstrating the mark's presence in the market.

Read More

Copyright and the third draft of the EU GPAI Code of Practice: towards a balanced governance of generative artificial intelligence

With the rise of General-Purpose Artificial Intelligence (GPAI) models, copyright law is reaching a critical crossroads. The European Union is now striving to strike a balance between technological innovation and safeguarding the fundamental rights of creators. In this context, it has proposed a Code of Practice on AI, aimed at framing the use of GPAI in alignment with the forthcoming European AI Act. Published in March 2025, the third draft of this Code is intended to be more operational and addresses the concerns raised by rights holders.

What is a GPAI Model?

A General-Purpose Artificial Intelligence (GPAI) model is an AI system designed to perform a wide range of tasks, such as drafting texts, translating languages, analysing legal documents, or generating images. It is not restricted to a specific use, making it highly versatile.

To achieve such performance, a GPAI model must be trained on massive volumes of data drawn from the internet, books, scientific articles, forums, images, and audio content. This training relies on machine learning techniques that enable the model to detect patterns within the data, understand language, and even mimic styles.

This ability to leverage a broad array of content, often protected by copyright, raises complex legal questions, particularly with regard to the lawfulness of the use of online content. As a result, the establishment of a framework, notably through the GPAI Code of Practice, is essential to support the development of these technologies while safeguarding the rights of creators.

Understanding the third draft of the GPAI Code

Key objectives and structure of the text

The third draft of the GPAI Code pursues several key objectives:

  • Enhancing transparency in the development and deployment of GPAI, through improved documentation of training data.
  • Framing the use of copyrighted content by encouraging developers to assess the legality of their datasets and comply with EU copyright exceptions.
  • Anticipating systemic risks by prompting providers to regularly evaluate the impacts of their GPAI models and adopt corrective measures.

The text adopts a modular structure, comprising general guidelines, documentation templates, technical recommendations, and frameworks for internal compliance policies.

Evolution from previous versions

The third draft represents a significant shift from earlier versions:

  • Clarified and streamlined commitments, better adapted to industry operators.
  • Standardised tools, such as “model cards,” to enhance transparency regarding the models’ features and datasets.
  • A more flexible approach, transitioning certain obligations from mandatory requirements to recommended best practices.

Copyright challenges in the age of artificial intelligence

The impact of GPAI models on protected works

The training of GPAI models relies on extensive datasets of textual, visual, and audio materials, a significant portion of which is protected by copyright. This gives rise to several major concerns :

  • Unauthorised use of protected works, without licenses or compliance with legal exceptions.
  • Creation of derivative content closely imitating existing works, without acknowledgment or compensation to the original authors.
  • Loss of authorship traceability, as models generate content disconnected from any identifiable source.

Legal grey areas

Several legal uncertainties remain:

  • Resorting to data mining (TDM) exceptions, particularly for commercial GPAI models, remains subject to debate.
  • Cross-border issues, especially where data is collected from jurisdictions with weaker copyright protections.
  • The opacity of training processes, which hampers the ability of rights holders to detect and address infringements.

How the third draft addresses copyright issues

Transparency and data traceability

The GPAI Code of Practice promotes a culture of accountability through documentation:

  • Standardised documentation templates enabling developers to describe the sources, types of data used, and legal justifications for their inclusion.
  • Recommended publication of data summaries, to enhance clarity and accessibility for regulators and third parties.

Commitments from GPAI Providers

The draft favours a flexible yet structured approach:

  • Publication of a copyright compliance policy: while optional, it is strongly encouraged.
  • Dialogue with rights holders: the Code suggests establishing reporting channels, without imposing an obligation to respond or act immediately.

Reception of the draft by stakeholders

The third draft has prompted contrasting reactions:

  • Some creators’ representatives view the draft as overly permissive, fearing a purely symbolic regulation without real enforcement mechanisms.
  • GPAI providers, on the other hand, welcome a more pragmatic and operationally realistic version, although they still call for greater legal certainty.

 

Conclusion

The third draft of the EU GPAI Code represents a significant milestone in the regulation of general-purpose artificial intelligence models, fostering a culture of transparency and encouraging respect for copyright. Nevertheless, by relying mainly on voluntary commitments, its impact on industry practices may remain limited without binding obligations.

 

Key Takeaways:

  • The third draft emphasises transparency through documentation, without imposing strict obligations.
  • The use of protected works remains loosely regulated, particularly for commercial GPAI.
  • A balanced approach between fostering innovation and protecting creators’ rights still needs to be achieved.

Need expert guidance on AI and intellectual property? Dreyfus Law Firm specializes in intellectual property law, including trademark, copyright, and AI-related legal matters.

 

We collaborate with a global network of intellectual property attorneys.

Join us on social media!

LinkedIn  

Instagram

Read More

Game of Thrones disqualifies Game of Döner from trademark registration

In a recent decision, the European Union Intellectual Property Office (EUIPO) ruled in favor of Home Box Office, Inc. (HBO), disqualifying the trademark application for Game of Döner on the grounds that it would unfairly benefit from the reputation of the well-known Game of Thrones trademark.

The EUIPO’s opposition division recently addressed a trademark dispute involving HBO’s Game of Thrones and a German applicant’s attempt to register Game of Döner for fast-food services.
The decision underscores the protection afforded to trademarks with a reputation under article 8(5) of the EU Trade Mark Regulation (EUTMR).

Background of the case

The applicant sought to register Game of Döner as an EU trademark for services in classes 35, 43, and 45.
HBO opposed the application, arguing that the use of Game of Döner would take unfair advantage of the reputation of its Game of Thrones trademark.

Legal framework: article 8(5) EUTMR

Article 8(5) EUTMR provides that a trademark shall not be registered if it is identical or similar to an earlier trademark with a reputation, and if the use of the later mark would take unfair advantage of, or be detrimental to, the distinctive character or repute of the earlier trademark.

Assessment by the opposition division

Reputation of the earlier mark

The opposition division acknowledged that Game of Thrones enjoys a high degree of reputation within the EU, particularly in relation to entertainment services.
HBO provided substantial evidence, including numerous awards and widespread merchandising, to support this claim.

Similarity of the signs

The signs Game of Thrones and Game of Döner were found to be visually and aurally similar to a certain degree.

Both share the distinctive phrase “Game of,” and the use of similar typography and imagery, such as a dragon emblem, further contributed to the perceived similarity.

Link between the marks

The division determined that the relevant public would likely establish a connection between the two marks, given the similarities and the reputation of Game of Thrones.
The use of a dragon in the Game of Döner logo was particularly noted as reinforcing this link.

Unfair advantage

It was concluded that the applicant would gain an unfair advantage by leveraging the reputation of Game of Thrones without making the necessary investments.

This would allow the applicant to benefit from the positive associations of the earlier mark, constituting an unfair advantage.

Conclusion

The EUIPO’s opposition division refused the registration of Game of Döner for all the classes it sought registration, citing the unfair advantage it would gain from the reputation of Game of Thrones.

Implications for trademark strategy

This case highlights the importance of conducting thorough trademark searches and assessments before filing applications, especially when the proposed mark bears similarities to well-known trademarks.
Businesses should be cautious to avoid potential conflicts that could lead to opposition proceedings and the refusal of trademark registration.

About Dreyfus Law Firm

At Dreyfus Law Firm, we specialize in intellectual property law, offering expert advice on trademark registration, opposition proceedings, and brand protection strategies.
Our partnership with a global network of IP attorneys ensures comprehensive support for our clients worldwide.

Dreyfus Law Firm is partnered with a global network of attorneys specializing in intellectual property law.

Join us on social media !

LinkedIn  

Instagram

FAQ

What is article 8(5) EUTMR?

Article 8(5) of the EU Trade Mark Regulation prevents the registration of a trademark that is identical or similar to an earlier trademark with a reputation if the use of the later mark would take unfair advantage of, or be detrimental to, the distinctive character or repute of the earlier trademark.

Why was Game of Döner refused registration?

The EUIPO found that Game of Döner would take unfair advantage of the reputation of Game of Thrones, leading to its refusal under article 8(5) EUTMR.

Does the presence of a dragon in the logo affect the decision?

Yes, the use of a dragon, a central element in Game of Thrones, in the Game of Döner logo reinforced the perceived link between the two marks.

Read More

Can AI be held liable for infringement or plagiarism?

In an era where artificial intelligence (AI) is deeply integrated into creative and decision-making processes, a key legal question arises: can AI be sued liable for infringement or plagiarism?
As AI systems rapidly generate texts, images, and music, understanding the legal framework is crucial for businesses seeking to secure their use of AI technologies.

This article explores the allocation of liability for intellectual property (IP) infringement involving AI, the challenges under current law, and strategies businesses can adopt to mitigate risks.

Legal liability of AI: Current framework

AI lacks legal personality

Currently, no legislation, including the recently adopted European Regulation 2024/1689 on Artificial Intelligence (AI Act), recognises AI systems as independent legal entities.
AI remains a tool, devoid of legal standing. Under article 1240 of the French Civil Code and general tort law principles, only natural persons and legal entities using, developing, or commercialising AI can be held liable for harm caused.

The AI Act outlines specific obligations for high-risk systems but emphasizes that compliance and accountability rest entirely with human or legal operators, not with the AI systems themselves.

Legal uncertainty for autonomous AI creations

Copyright law, particularly under the Berne Convention and Directive 2001/29/EC, protects works created by human authors.

AI-generated content created independently, without substantial human intervention, falls outside traditional legal protections, making ownership claims and infringement actions more complex, particularly when such content unlawfully reproduces pre-existing works.

The AI Act does not create an independent IP regime for AI-generated works. However, Article 50 mandates transparency obligations, requiring users to be informed when interacting with artificially generated content. At the same time, Article 52 establishes procedures for overseeing AI models that pose systemic risks.

When AI-generated content leads to infringement: who can be held liable?

Liability of natural and legal persons operating AI

When AI-generated content infringes third-party rights, the operator of the AI system is treated as the legal author of the infringing content.

Liability is incurred regardless of intent, as exploitation alone can trigger legal consequences if it causes harm to a rights holder.

To reduce such risks, operators must:

  • Conduct systematic prior verification of AI-generated outputs;
  • Implement internal compliance protocols tailored to the specific AI system.

Liability of AI developers and suppliers

AI developers or suppliers, whether natural persons or corporate entities, can also be held liable in two key situations:

  • Failure to ensure effective human oversight, including by not providing users with the necessary information for proper understanding, monitoring, and intervention, as required by Article 14 of the AI Act;
  • Unlawful use of protected works during the model training process, constituting a separate IP infringement.

In accordance with the AI Act, providers of high-risk AI systems must:

  • Ensure the quality, representativeness, and statistical relevance of training, validation, and testing datasets (Article 10);
  • Provide detailed technical documentation, describing system characteristics, development processes, and compliance measures (Article 11);
  • Inform end users about the nature and limitations of AI-generated content, ensuring transparency and protecting user expectations (Article 50).

The DeepSeek case: a cautionary tale of AI plagiarism

In March 2024, Chinese company DeepSeek was accused of plagiarising large portions of copyright content without proper attribution.
Investigations revealed that the AI chatbot reproduced entire sections word-for-word, lacking substantial transformation or originality.

DeepSeek defended itself by arguing that the sources were public and that the reuse constituted transformative use.

However, according to international copyright standards, mere aggregation or superficial rewording is insufficient to avoid infringement when the original content remains recognisable.

This case highlights the substantial risks businesses face when training datasets are not properly vetted. It stresses the importance of:

  • Rigorous auditing of training datasets;
  • Maintaining detailed records of dataset provenance;
  • Implementing robust transparency measures for AI-generated content.

Strategies to protect your business

Establish strong internal compliance policies

  • Systematically review AI outputs before publication;
  • Train employees on intellectual property risks associated with emerging technologies;
  • Limit AI use to cases where legal risks are effectively managed.

Strengthen contractual safeguards

  • Request explicit warranties regarding the legality of training data;
  • Negotiate robust indemnification clauses to cover potential IP infringements;
  • Reject unreasonable limitations of liability in supplier agreements.

Deploy technological and legal risk management tools

  • Use advanced plagiarism detection software tailored for AI-generated content;
  • Establish internal auditing procedures and rapid response mechanisms to address identified risks.

Conclusion: anticipating and managing legal risks

AI systems, under current law, cannot be held liable.

Legal responsibility for infringement or plagiarism always rests with the natural or legal persons developing, operating, or commercialising AI solutions.

Given the new regulatory obligations imposed by the European AI Act, companies must adopt a proactive approach, integrating technical diligence, contractual safeguards, and ongoing legal oversight.

Dreyfus Law Firm works with clients in the food sector, providing specialist advice on intellectual property and regulatory issues to ensure compliance with national and European laws.

We collaborate with a global network of intellectual property attorneys.

Join us on social media !

FAQ

Can AI be sued for plagiarism or infringement?

No. Only a natural or legal person operating or developing an AI system can be held liable.

How can a company minimise its legal risks when using AI?

By implementing strict validation procedures, securing contractual protections, and regularly auditing AI-generated outputs.

Does the European AI regulation impose obligations relating to intellectual property?

Indirectly, through transparency and dataset quality requirements applicable to AI developers and deployers.

 

Read More

The social media threat: a fashion and luxury perspective

Digital infringement is no longer confined to marketplaces or anonymous domains. Today’s threatscape is defined by a structural shift in both consumer behaviour and vendor strategy: from e-commerce websites to media-first platforms, direct-to-consumer messaging apps, and short-form content ecosystems.

Social commerce is not only accelerating in reach, but fragmenting enforcement efforts, placing unprecedented pressure on brand owners to adapt, monitor, and act swiftly across decentralized channels.

Main threats targeting luxury brands on social media

Impersonation accounts and fake profiles

An alarming 22 million users are following fake fashion and luxury brand accounts across platforms. The average number of followers for counterfeit profiles has surged by 20% in 2024 alone, compared to late 2023.

These accounts typically rely on:

  • Full impersonation techniques using stolen imagery and layouts
  • Brand handle spoofing: e.g., “brandname.official” or regionalized variants
  • Bot-controlled engagement farms that artificially inflate credibility

Multi-channel counterfeit networks

Sophisticated counterfeit operators utilise multi-channel ecosystems, combining Instagram, TikTok, WhatsApp, and external fake shops to:

  • Publish “zombie listings”
  • Direct consumers to covert sales channels via private messages
  • Avoid detection by removing sales elements from visible platforms

These structures create encrypted, evasive infrastructures, requiring advanced monitoring tools and legal foresight.

Ephemeral content tactics

The exploitation of disappearing formats like Instagram Stories or Snapchat posts—which vanish after 24 hours—greatly complicates detection. These are used to:

  • Launch flash sales of counterfeit goods
  • Evade traditional traceability
  • Engage in coordinated bot repost attacks (botnets) with minimal digital footprint

Paid advertising exploitation

Malicious actors purchase social ads to actively promote counterfeit goods. Paid ads, once a trusted visibility tool, now contribute to the global reach of fake listings.

Without robust monitoring, these campaigns risk legitimising fakes in the eyes of unaware consumers.

Influencer-fueled “dupe” culture

A growing number of influencers collaborate—intentionally or unknowingly—with counterfeit promoters via affiliate schemes. In exchange for commissions, they help distribute illicit goods to mass audiences under the banner of “affordable luxury alternatives.”

This trend normalises counterfeiting, challenges brand integrity, and capitalises on viral attention cycles.

A strategic pillar-based brand protection framework

Monitoring and enforcement

An effective strategy starts with granular surveillance and prioritised action:

  • Identify high-impact infringers
  • Remove paid ads and deceptive reviews
  • Set thresholds for enforcement to optimise legal spend
  • Monitor seller communications and platform evolution
  • Quantify revenue loss and develop restitution pathways

Platform engagement and proactive escalation

Brands must shift from passive complaints to structured engagement with social platforms:

  • Audit their enforcement policies and escalation timelines
  • Supply real-time keyword lists and high-risk imagery
  • Negotiate recurring sync meetings with internal compliance teams

Rights-holder and institutional collaboration

Brand protection becomes exponentially more effective through:

  • Joint actions with IP agencies and national authorities
  • Peer-to-peer intelligence sharing among industry stakeholders
  • Cross-functional teams linking IP, cybersecurity, marketing, and legal

Key metrics to assess the brand protection

Quantifying value recovered from protection programmes is vital. Indicators include:

  • Revenue loss per engagement: e.g. Post reach × $0.05 cost per engagement (industry benchmark)
  • Illicit listing value:
    Listings removed × Avg. stock × Avg. price × 60% sell rate × 40% conversion rate × Gross margin
  • Brand sentiment & reputation:
    Social sentiment scores, complaint incident data, retailer feedback
  • Consumer trust:
    Blockchain authenticator interactions, re-purchase rates, refund/litigation trends

Conclusion: elevating protection, restoring trust

The luxury industry faces a crossroads: protect aggressively or concede ground to digital counterfeiters. With data-driven enforcement, platform engagement, and stakeholder alliances, fashion brands can turn vulnerability into opportunity, strengthening both revenue and consumer loyalty.

Dreyfus Law Firm offers tailored brand protection strategies rooted in IP law, with a proactive and international scope.

Dreyfus Law Firm is in partnership with a global network of Intellectual Property law specialists.

Join us on social media !

LinkedIn  

Instagram

FAQ

1. How many users follow fake fashion brand accounts?

More than 22 million globally, with a 20% increase in follower count in 2024 alone.

2. Why is ephemeral content hard to police?

It disappears within 24 hours, leaving little evidence for enforcement or litigation.

3. Are bots still a significant threat?

Yes. From repost storms to hashtag hijacking, bots remain central to counterfeit amplification.

Read More