Employers beware: CO2 reporting of employee’s travel mandatory as of 1 July 2024

From 1 July 2024, employers with more than 100 employees will be obliged to record and report their employees’ commuting and business travel. This ‘reporting obligation work-related personal mobility’ stems from the Climate Agreement.

What does the reporting requirement entail?

The reporting requirement means that an employer must report kilometres travelled by its employees. There is no requirement to track actual CO-2 emissions.

The reporting obligation distinguishes between commuting and business travel. Commuting involves trips made by employees between their place of residence or stay and the location where work is performed. Business travel includes the trips an employee makes for work, excluding commuting. Business travel is broken down into: Lease and/or own fleet, mobility services and expense claims.

Furthermore, kilometres must be reported split by fuel type and means of transport. In terms of means of transport, cars, motorbikes, mopeds/scooters, (e-)bicycles, walking and public transport are covered by the reporting obligation. Other means of transport are excluded from the reporting obligation.

Who counts as an employee?

For the calculation of the 100 employees, all employees who work more than 20 hours per month under an employment agreement or on-call agreement count. Temporary workers, seasonal workers, self-employed workers and volunteers do not count in the calculation. For temporary employment agencies, temporary workers do count unless an temporary agency clause is included in the employment agreement.

When to report?

Employers employing more than 100 employees on 1 July 2024 must report employee business and commuting traffic for the second half of 2024 (or all of 2024) by 30 June 2025. Employers that hire more than 100 employees after 1 July 2024 will not yet have to report for 2024 in 2025. It is of course also possible to report voluntarily.

In conclusion

More information can be found in the Ministry of Infrastructure and Water Management’s handout. You can do the reporting yourself on the website of the Rijksdienst voor Ondernemend Nederland. If you have any questions, you can of course contact us.

From friend to foe: what you need to know about the law on adjusting dispute resolution and clarifying admissibility requirements inquiry procedure (Wet aanpassing geschillenregeling en verduidelijking ontvankelijkheidseisen enquêteprocedure)

Imagine this: you start a company full of enthusiasm and confidence with a few close friends or business partners. You share the same dreams and goals. But what happens when you suddenly disagree?

Ending shareholder disputes

If you cannot manage to work things out together, then ending the shareholder dispute can be a huge challenge. Current legislation allows for the squeeze-out of another shareholder (uitstoting) or withdrawal (uittreding) (the “dispute resolution”) in circumstances, but the court procedures for doing so are not currently perceived as accessible or practical. They cost a lot of time and money. For the time being, the possibility of having an enquiry into the affairs and policies of a legal entity (the “inquiry procedure”, de enquêteprocedure) is more frequently used, as it is generally perceived as somewhat more accessible and effective. However, this procedure also involves costs and takes time.

The law on adjusting dispute resolution and clarifying admissibility requirements inquiry procedure (Wet Aanpassing geschillenregeling en verduidelijking ontvankelijkheidseisen enquêteprocedure)

The “Wet Aanpassing geschillenregeling en verduidelijking ontvankelijkheidseisen enquêteprocedure” (also called “Wagevoe”) is designed to solve disputes within legal entities faster and more efficiently:

  • Thus, the Enterprise Chamber becomes the body to conduct both dispute resolution and inquiry proceedings. Concentrating proceedings at the Enterprise Chamber is expected to save time and money.
  • Whereas previously a squeeze-out/withdrawal procedure had to be started with a writ of summons and an inquiry procedure with a petition, all the aforementioned procedures will now be started with a petition. This makes it possible to combine the litigation procedure with the inquiry procedure. This is expected to save time and money.
  • It will also (presumably) become easier to squeeze-out a shareholder: previously, the courts only looked at the conduct of the “obstructive” shareholder as such, but with the introduction of the new law, other conduct, for example his conduct as a director, will also be considered. This may make the squeeze-out regime more attractive for resolving disputes between, for example, two shareholders also directors, where there is a deadlock not only in the general meeting, but also in the board.
  • In addition to shareholders, certificate holders will now also have the possibility to initiate an exit procedure. Especially for certificate holders whose position is similar to that of a shareholder, this seems to be a favourable instrument.
  • Also, the access requirements for the inquiry procedure with regard to capital providers of listed companies with an issued share capital not exceeding €22.5 million will be clarified: a capital provider will in principle have access to the inquiry procedure if it represents 1% or more of the issued capital or represents a value of at least €20 million.

Prevent escalation

Although the arrival of the new law is expected to make dispute resolution and the inquiry procedure more attractive, it is preferable to stay in control yourself. After all, litigation costs time and money. By drawing up your own arrangement, you can avoid ending up in court. After all, you will have a script ready for when shareholders get into conflict with each other. But even if you do end up in court, having your own arrangement is useful. After all, your preferred and tailor-made arrangement usually takes precedence over the legal dispute resolution scheme, unless a share transfer on the basis of that own scheme is not possible or extremely objectionable. So it is definitely worth making one.

Possibilities

For example, include your own arrangement in the articles of association and/or make a shareholder agreement stating what happens if there is a deadlock situation. For example, you can stipulate that each of the two shareholders in a deadlock situation has the right to initiate a bidding procedure in which one shareholder (X) has to offer his shares to the other shareholder (Y), in which the shareholder Y has the choice of taking over the shares or transferring his own shares to shareholder X for that price (“Russian roulette clause”) or that both shareholders make a closed bid for each other’s shares where the highest bidder is obliged to take over the other shareholder’s shares (as, for example, in the “Texas shoot-out clause” or the “Mexican shoot out clausule”).


There are numerous possibilities to make your own arrangement. Wieringa Advocaten will be happy to advise you on this and can draft a tailor-made arrangement for you. Feel free to contact us.


Are you already involved in a (shareholder) dispute? We will be happy to advise you and, if necessary, assist you in court.

Twenty-year anniversary of Wieringa weblog!

Exactly 20 years ago today, we published our first blog post on www.wieringa.nl. That first blog post (in those days also referred to as a blawg) was written by Lex Bruinhof and – coincidentally – was about the scope of the intellectual property of sponsor brands during UEFA Euro 2004 in Portugal.

In the following twenty years, we published over 4,500 blogs. That way, we keep our clients, other interested parties and, of course, ourselves informed about recent legal developments in a pleasant way. The 20th anniversary of our weblog is a good opportunity to make our blogs available to even more people in the future. Therefore, from now on our blogs are also available in English! In other words, blawgs again.

If you have a question about a blog, please feel free to contact us.

Everyone is allowed to sell Big Macs from now on! Or aren’t they?…

Last week, I saw an article of nu.nl in my LinkedIn feed with a title that caught my attention:“Anyone can sell a Big Mac (with chicken)“. When I then Googled I found several more newspaper articles with similar headlines and text. For example, NRC writes:“Anyone may now sell a ‘Big Mac’ if it is a sandwich with chicken“. And the AD writes: “Any burger baker may use the name Big Mac“. In case reading these headlines made you think “that sounds like a gap in the market!”, I advise you to read on quickly.

The newspapers seemingly draw this conclusion based on a recent European Court of Justice ruling. In it, the court ruled that the European Union Intellectual Property Office (EUIPO) was right to declare the trademark “Big Mac” lapsed for food prepared with poultry and sandwiches with chicken. However, how the various newspapers came to the conclusion on that basis that any burger baker can now sell his (chicken) burgers under the “Big Mac” brand is mystery to me. To prevent a lot of burger bakers from receiving a writ of summons from Ronald Macdonald in the near future, I think it would be good to dispel this myth of the quality newspapers in this blog.

The Big Mac trademark

On 1 April 1996, McDonalds filed the trademark “Big Mac” as a European Union wordmark. And whether you love or hate it, almost everyone will be familiar with the famous burger and its trademark. There is even an actual (somewhat playfully intended) economic index named after it – the Big Mac index – which was conceived by The Economist in 1986. After all, the Big Mac is sold almost everywhere in the world and contains roughly the same simple ingredients everywhere. A very strong trademark that many trademark owners can only dream of. Perhaps the trademark is almost as strong as the trademark owner’s own name registered as a trademark.

Goods and services

A trademark is always registered for specific goods and services. Thus, this also applies to the trademark “Big Mac”, which was registered in 1996 for, among other things, foods prepared with meat in general, but also for foods prepared with specific categories of meat, namely pork, fish and poultry. In addition, the trademark was specifically registered for sandwiches containing meat, fish, pork and chicken. Ironically, not specifically for sandwiches containing beef; while in practice the Big Mac is prepared with beef (India aside, but the burger has a different name there). The trademark was additionally registered for a wide range of other ingredients, including preserved and cooked fruits and vegetables, eggs, cheese, milk, milk preparations, pickles, desserts, biscuits, biscuits, biscuits, chocolate, coffee, coffee substitutes, tea, mustard, oatmeal, pastries, sauces, seasonings and sugar.

Declaration of revocation

There is an obligation on trademark owners to put a registered trademark to genuine use in relation to the goods and services for which it is registered. If a trademark owner fails to do so for a continuous period of five years, it risks revocation its trademark under Article 58(1)(a) of the European Union Trade Mark Regulation. Where a trademark owner takes action against an alleged infringer, the alleged infringer has the option of filing a counterclaim for revocation.

And this is what happened in 2017. An Irish burger restaurant called Supermac’ s got caught up in proceedings in which McDonalds invoked its “Big Mac” trademark. Supermac’s responded by filing a claim for revocation of the Big Mac mark. This was because, according to Supermac’s, McDonalds was using its trademark exclusively for the famous beef sandwich – which was not included in the trademark registration. The EUIPO went along with that argument and declared the revocation of the “Big Mac” trademark.

Fourth Board of Appeal

McDonalds appealed, arguing first that beef and chicken fall under the broader categories of meat and poultry. Moreover, it would have actually sold burgers containing both beef and chicken in the past five years. It presented very substantial evidence for this. To substantiate the sale of chicken burgers, it recited that, among other things, it allegedly sold a Big Mac in France that was prepared with chicken meat, as allegedly shown in the Facebook screenshot below.

The Fourth Board of Appeal ruled on 14 December 2022 that chicken meat does indeed fall under the broad category of poultry. Moreover, the evidence submitted would show that the trademark was indeed normally used for sandwiches prepared with beef and chicken meat. The decision to declare the trademark’s complete revocation was therefore reversed.

Normal use for sandwiches containing pork and fish, as well as all the other ingredients mentioned earlier, was not accepted by the Board of Appeal. While it is true that some ingredients were used in the preparation of the Big Mac, the Board of Appeal did not find it plausible that the public actually associates the trademark “Big Mac” with these ingredients. After all, no one talks about Big Mac vegetables, Big Mac cheese, or Big Mac pickles.

Court of Justice

The case eventually ended up in the Court of Justice. There, Supermac’s argued that the evidence previously submitted by McDonalds as far as the Big Mac with chicken was inaccurate. First of all, the advertising material supplied would only consist of drafts – as evidenced by the text “confidential” included therein. Drafts of advertising material obviously do not conclusively prove that a trademark is genuinely used in practice. Moreover, the screenshots provided by Facebook (including as shown above) would date from 2016; and thus fall outside the required five-year period. This, according to Supermac’s, did not prove that McDonalds had used its trademark for poultry and chicken sandwiches in the past five years. The ECJ agreed and rejected these goods.

Consistency

The consequence of the revocation is that the trademark Big Mac is no longer registered for food prepared with poultry and sandwiches with chicken. And that’s all there is to say about it. The fact that the trademark is no longer registered for food prepared specifically with poultry and specifically for chicken sandwiches does not mean that the trademark can therefore be freely used by anyone wishing to sell chicken burgers under the “Big Mac” trademark.

First of all, beacause the trademark is still registered for foods prepared with meat in general. When a court considers whether there is a likelihood of confusion, it will still take into account that meat and chicken are complementary to each other. The greater simply includes the lesser. The trademark registration may have become less specific, but that does not mean that the trademark has thereby lost its protection entirely in respect of the more specific goods that have been crossed out.

Trademark with a reputation

Moreover, McDonalds has another important legal ground to act against parties using the “Big Mac” trademark (or trademarks that are similar). The trademark is considered to be a trademark with a reputation. And such trademarks enjoy a very broad protection under Article 9(2)(c) of the European Union Trade Mark Regulation. In fact it does not even matterr for which goods or services an identical or similar trademark is used. As long as this causes confusion and where the use without due cause of the trademark would take unfair advantage, or be detrimental to, the distinctive character or the repute of the trademark. There is no doubt that this would be the case if anyone other than McDonalds were to use the “Big Mac” trademark for other goods or services. Be it chicken burgers or flip-flops. The Dutch newspapers quoted earlier have drawn a conclusion that (at least in terms of trademark law) is not based on anything.

Conclusion

Were you planning to start selling Big Macs prepared with chicken tomorrow? If so, you would be wise to reconsider your business plan. There could be an angry clown at your doorstep in no time… Of course, you can then still try to recover your losses from one of the quoted Dutch newspapers.

Employment law in the Netherlands – part 3 key features of Dutch Labor Law

The Dutch labor law framework is characterized by a balance between employee protection and entrepreneurial flexibility, rooted in a tradition of social dialogue and progressive regulation. Understanding the key features of Dutch labor law is crucial for both employers and employees operating within the Netherlands. In this third part of our comprehensive guide into Dutch labor laws, we highlight some essential elements that define this legal landscape.

1. Employment Contracts

Dutch law recognizes three main types of employment contracts:

  • Permanent contracts: these are open-ended contracts that provide significant job security. Termination is subject to stringent legal requirements, such as prior court proceedings.
  • Fixed term Contracts: These are for a specified period and can be renewed up to three times within a three-year period before they automatically convert into a permanent contract.
  • Temporary Agency Contracts: These involve a tripartite relationship between the employee, the temporary employment agency, and the company where the employee actually works.

Each type of contract must comply with specific legal requirements regarding terms, conditions, and notice periods. Other contract types exist as well, such as payroll contracts.

2. Probationary Period

Probationary periods in the Netherlands are strictly regulated. For permanent contracts and fixed-term contracts longer than two years, a probationary period of up to two months is allowed. For fixed term contracts with a duration between six months and two years, only one month is allowed. Fixed term contracts shorter than six months, a probationary period is not permitted. During the probationary period, both employer and employee can terminate the contract with immediate effect for any reason other than discriminatory grounds.

3. Working hours and overtime

The standard workweek in the Netherlands is 40 hours, but variations exist depending on the sector and specific agreements. The Working Hours Act (Arbeidstijdenwet) governs maximum working hours, rest periods, and overtime:

  • Maximum hours: Employees can work a maximum of 12 hours per shift and 60 hours per week, though these limits are subject to averaging over a given period.
  • Overtime: Compensation for overtime is typically agreed upon in the employment contract or collective labor agreements (CLAs). There is no statutory requirement for overtime pay, but – at least in some sectors – it is regularly provided for.

4. Leave and holidays

Employees in the Netherlands are entitled to various types of leave:

  • Annual leave: The statutory minimum is four times the weekly working hours (e.g., 20 days for a full-time employee). Additional days are often provided by CLAs. Best practice as well shows employer offering 5 above statutory holidays on a yearly basis.
  • Public Holidays: There are generally eight public holidays, though the exact number can vary.
  • Sick Leave: Employers must pay at least 70% of the employee’s salary for up to two years of illness. The salary does not only include base wage, but potentially as well commission or bonus.
  • Parental Leave: Parents are entitled to parental leave, which can be taken up until the child reaches eight years of age.
  • paternity leave: relatively recently additional rights have been introduced, providing more support for new fathers. Now 5 additional weeks are offered to be taken within 6 months after the child is born. The employee’s wage is covered by the UWV up to 70%.

Other types of leave exist as well, such as pregnancy leave and short term and long term care leave.

5. Dismissal and Termination

Dutch law imposes strict rules on dismissals to protect employees from unfair termination:

  • Dismissal with permission: employers need prior permission from the Employee Insurance Agency (UWV) or a court to terminate an employment contract.
  • Grounds for dismissal: before permission is granted, valid reasons for termination need to be proven. Valid reason may be economic reasons, long-term incapacity, poor performance, and misconduct.
  • Notice period: the notice period varies based on the length of service, generally ranging from one to four months. CLA might deviate from this, just as employment agreements.

In cases of unfair dismissal, employees can claim compensation or reinstatement. Additionally, when the employer terminates the employee’s employment, in principle the employee is entitled to a statutory severance pay (transitievergoeding), being 1/3 of a monthly wage (including holiday allowance and commission and/or bonus) per year worked.

6. Collective Labor Agreements (CLAs)

Many industries and sectors in the Netherlands are governed by collective labor agreements, which set terms and conditions of employment, including wages, working hours, and benefits. These agreements are negotiated between employers’ organizations and trade unions and are legally binding if employers are a member of the employers organizations, if the CLA is incorporated into the employment agreement and/or if the CLA has been declared universally applicable by the Dutch government.

7. Employee representation

Dutch law promotes strong employee representation through Works Councils (Ondernemingsraad):

  • Works councils: companies with 50 or more employees are required to establish a Works Council. These councils have various rights, including the right to be informed and consulted on significant business decisions. As well they have a right of consent on several topics, such as the implementation of reward policies.
  • Representation in smaller companies: In smaller companies with between 10 and 50 employees, employee representatives may be elected to ensure workers’ voices are heard. When the employees request for such a representative, the company is obliged to appoint.

Conclusion

Dutch labor law provides a robust framework that balances the interests of employees and employers. Its key features, from comprehensive employment contracts and strict dismissal protections to detailed regulations on working hours and leave entitlements, make it one of the most protective labor systems in Europe. Understanding these laws is essential for maintaining compliant and fair workplace practices in the Netherlands.

Want to learn more e.g. through an in-house training? We are happy to provide.

The non-competetion clause – Special clauses in the employment agreement

An employment agreement sets out essential terms such as the position, salary and holiday and sick leave rights. In addition, provisions are often included on, for example, probationary periods and joining a competitor after termination of the employment relationship. Separate different requirements apply to these ‘special clauses’, both legally and developed in case law. In this series, we discuss the most common special clauses. In this episode: the non-competition clause.

Background

The non-competition clause restricts an employee after termination of employment from performing work that competes with the former employer’s business. The purpose of the non-competition clause is to protect an employer’s market position. After all, an employee may use business-sensitive information at a competitor, potentially causing damages tothe former employer. The clause is explicitly not meant to bind employee. At play here is the fact that restricting work is contrary to the fundamental right to free choice of employment. For this reason, the non-competition clause is subject to strict requirements.

Non-competition clause requirements

First, a non-competition clause must be included in writing in the employment agreement. It is also sufficient that an employment agreement or letter refers to the non-competition clause in attached terms and conditions of employment, as long as an employee has signed the employment agreementor letter.

Second, a non-competition clause is only valid in an employment agreement for an indefinite period of time. The non-competition clause in a fixed-term contract is null and void in principle, unless an employer motivates that the clause is necessary because of important business interests. An employer must justify which business interests justify use of the clause, but also why the non-competition clause is necessary for the specific employee.

Furthermore, both indefinite and fixed-term agreements should specify the duration of the restriction and the geographical scope of the non-competition clause. Although this is not a legal requirement, case law shows that violation of this can be grounds to annul the clause in whole or in part.

Finally, the non-competition clause must be re-agreed if the non-competition clause has become significantly more onerous because of a (major) job change. This may be the case if an employee has been promoted to a significantly higher position. An employee must then again be given the opportunity to properly consider the consequences of the clause.

Validly agreed, now what?

If a non-competition clause is validly agreed in an employment agreement, an employer can claim that an employee complies with the non-competition clause. This is regularly accompanied by a penalty clause in the employment contract that serves as an incentive not to breach the clause. Conversely, an employee can claim (partial) annulment or suspension of the clause if he is unfairly disadvantaged by use of the non-competition clause. This is the case if an employee’s interest in being employed elsewhere, given the circumstances, outweighs the employer’s interest in binding the employee to the non-competition clause. Missing out on a higher salary elsewhere or better career prospects might be a factor in this. To accommodate each other’s mutual interests, parties can also agree to different arrangements regarding the non-competition clause in a settlement agreement.

Possible new requirements

Relatively recently, Social Affairs and Employment Minister van Gennip expressed her intention to tighten the requirements of the non-competition clause. Proposals include a limitation of duration and introduction of a fixed penalty. We previously wrote extensively about this bill.

Conclusion

We regularly help entrepreneurs and employees draft and assess non-competition clauses. Are you in doubt (following the above) about the validity and effect of your non-competition clause? Feel free to contact us, it is our pleasure to assist you.

Directors’ liability for tax debts

This blog is part of the Directors’ Liability Series.

Company directors bear a great responsibility, not only towards employees and shareholders, but also towards the tax authorities. An often underestimated aspect of this responsibility is the potential personal liability for tax debts of the company. This is because the tax authorities have far-reaching powers to hold directors personally liable if the company fails to meet its tax obligations. In this blog, we explain what the risks are and how directors can limit this liability.

Types of taxes

Under Section 36 of the Tax Collection Act 1990, a director of a company (an “entity” within the meaning of Section 2(1)(b) of the General Act on State Taxes) can be held jointly and severally liable by the tax authorities for the following taxes:

  • Payroll tax: The tax that the company must withhold from employees’ wages and remit to the tax authorities.
  • Sales tax (VAT): The tax that the company has to pay on the goods and services it supplies.
  • Corporate tax: The tax on the company’s profits.
  • Excise duties: Tax on certain products such as alcohol and tobacco.

Cases of liability – Notification of inability to pay

If the company is unable to pay the tax, the company is obliged to notify the Inland Revenue within two weeks of the due date of the tax assessment. Each director is authorised to fulfil this obligation on behalf of the company. Depending on whether such notification has been made (on time), the following situations apply:

  • Situation 1: timely reporting: if the company has complied with its obligation to report the inability to pay in due time, a director is liable if it is plausible that the non-payment of the tax debt is the result of manifestly improper management attributable to him in the period of three years preceding the time of the notification. The burden of proof for mismanagement lies with the tax authorities.
  • Situation 2: no (timely) notification: if the company has not (timely) complied with its obligation to report the inability to pay on time, in the context of the director’s liability, it will be presumed that the non-payment is due to the director and the three-year period will be deemed to start at the time the entity is in default. That presumption can only be rebutted if the director makes it plausible that it was not his fault that the body failed to fulfil its reporting obligation. Thus, the reverse burden of proof in this case lies with the director.

Liable persons

Liability for tax debts applies not only to the current director(s), but to former directors during whose administration the tax debt arose, de facto policymakers and, if the director is a legal entity, the directors of that legal entity.

Conclusion

As a company director, it is crucial to be aware of your obligations towards the tax authorities and the risks of personal liability. By timely notification of payment default, careful governance and seeking professional advice, you can avoid many problems. If you have any questions on this topic or need help taking the right steps, please feel free to contact us. Our team of experienced lawyers is ready to support and advise you.

In this blog, we identify a directors’ liability risk that has its origins in tax law. Wieringa Advocaten does not provide tax advice.

Questions to the CJEU: A private copy levy due for offline streaming?

Section 16b and 16c of the Dutch Copyright Act (Auteurswet) provide that one may make a few copies of copyrighted works for one’s own practice, study or use, as long as such use does not have a commercial purpose. A copy for one’s own use is called a “home copy” (thuiskopie). If all conditions are met then no permission from the copyright holder is needed for that copy. If that copy is made on a physical medium, such as an external hard disk, phone or tablet, then a small fee is owed to the copyright holder, the so-called private copy levy. That fee is factored into the purchase price of these physical mediums. So without realising it, you have probably already paid a private copy levy many times.

Offline streaming

Several streaming services now offer the option of “offline” streaming. This means that music, a film or series is temporarily downloaded onto a device so that it can be listened to or watched offline at another time. Usually, streaming services impose restrictions on this. For example, a film or series is often available offline for up to 30 days after downloading it. Of course, when cancelling a subscription, the ability to listen to or watch previously downloaded music or films also logically expires.

HP v Stichting Onderhandelingen Thuiskopievergoeding

HP took the position in 2019 in proceedings against the Stichting Onderhandelingen Thuiskopievergoeding because of the aforementioned restrictions that offline copies of music and films, as provided by streaming services, do not qualify as home copies. In HP’s view, streaming services control those home copies and not the end users. Of course, HP’s position is not surprising for a laptop manufacturer. To determine the amount of the home copy levy of data carriers is how often home copies are made on a given physical carrier.

The court of The Hague ruled in 2019 that offline streaming copies (also known as “tethered downloads“) fall to be classified as home copies within the meaning of the Copyright Act. Indeed, according to the court, they were in fact reproductions made for private use. The fact that people already pay a licence fee through streaming services as well did not, according to the court, detract from the Dutch system of home copying fees. The court also found it unnecessary to ask preliminary questions about this to the ECJ.

Court of Appeal of The Hague

In 2022, the case was again submitted to the Hague Court of Appeal. HP took the position that a tethered download is not a home copy because it is not made by a natural person and for his own use. Technically, this is indeed the case. After all, the copy is not made by the user but by the streaming service. The user only has to press the “download” button, effectively giving an instruction to the streaming service. However, Section 16b of the Copyright Act – which deals with copies of music and digital copies – does not require someone to make a copy himself. One may also instruct someone else to do so.

However, the court of appeal noted that it can be deduced from the parliamentary history that a copy does lose its status of a private copy when the other party acts with a commercial purpose (read: the streaming service). The Court of Appeal therefore set aside the earlier judgment by the court of The Hague and ruled that no equitable compensation is due for offline streaming copies(tethered downloads).

Questions to the Court of Justice

The case was referred in cassation to the Supreme Court, which decided on 17 May to refer the following preliminary questions to the ECJ. It is therefore up to the ECJ to tie the knot on this issue.

1. Can a reproduction with the characteristics described above in 2.3 at (i)-(iii) (offline streaming copy or tethered download), having regard also to the three-step test (Art. 5(5) Copyright Directive), be classified as a “reproduction (…) made by a natural person for private use, and without any direct or indirect commercial purpose” within the meaning of Art. 5(2)(b) Copyright Directive?

2. Is a national regulation under which the home copying exception does not include offline streaming copies consistent with the objectives of the Copyright Directive, including a high level of copyright protection, a fair balance between the interests of the rightholder and the interests of the user, and a coherent and technologically neutral application by Member States of the exceptions and limitations?

We are monitoring this case and will keep you informed of further developments!

Combating deepfake porn under civil law

One Dutch celebrity after another recently announced that they had fallen victim to deep fake porn videos. These are AI-edited videos in which victims ostensibly feature, while in reality this is not the case. This has major consequences for them. Recently, several victims announced that they will collectively file criminal charges. A criminal charge can lead to the conviction of creators (and sometimes those who publish the video on their own, i.e. publicists) of deepfake porn videos. Unfortunately, a criminal conviction does not undo the online publication of the videos. Like an oil slick, images continue to spread online and the disclosers are often based abroad. Publishing the footage is not always punishable abroad. In this blog, we look at what civil law remedies are available to victims to fight online publications of the footage.

Deepfake porn?

AI software makes it possible to fit one person’s face over another person’s face, as it were, in existing moving image material. The algorithm does that all by itself so there is no manual editing involved. So it is child’s play. The face edited into the footage even adopts the facial expressions of the original face. This now looks lifelike and has major implications for victims.

Criminal route

Several Dutch celebrities announced recently that they will file criminal charges. Those criminal charges are likely to focus mainly on the makers of the videos. Last year, for the first time in the Netherlands, a creator of deepfake porn was sentenced to 180 hours of community service. That conviction was based on Article 139h of the Penal Code, which criminalises so-called “revenge porn”. This centres on the manufacture of an image of a person of a sexual nature. Although the deepfake videos did not contain a real image of a person, the Amsterdam court ruled that deepfake sexual imagery could also qualify as an image of a sexual nature. This paved the way for the criminal conviction of creators of deepfake porn.

Civil law bases

The problem with deepfake porn videos is that their creator will be unknown in most cases. Of course, for a criminal conviction, the creator must first be tracked down. Moreover, a criminal prosecution does not always lead to a conviction.

Although publicists could also be criminally prosecuted under Dutch law, to my knowledge this has not yet happened in practice. In addition, the criminal prohibition of revenge porn is not regulated by law in every country. So the criminal route does not always offer a solution for victims.

Making and publishing the footage is not only criminally culpable. Publishing such visual material can also be unlawful in a civil law sense. This therefore gives victims another option besides criminal law to act against the publication of deepfake porn videos.

Portrait rights

The first way to tackle such videos civilly lies in portrait law, which is part of the Copyright Act. Portrait law allows a person portrayed to oppose the publication of their portrait under circumstances. The Copyright Act distinguishes between commissioned and non-commissioned portraits. In the case of deep-fake porn videos, these are (obviously) not commissioned portraits. The publication of such a portrait is not allowed if the person portrayed has a reasonable interest against it. This therefore involves a weighing up of interests.

A reasonable interest against disclosure may consist first of all of a commercial interest. For example, BN celebrities may oppose the disclosure of their portrait if they have a so-called redeemable popularity. The rationale behind this is that others should not profit from someone’s accumulated reputation without permission.

A so-called moral interest is also considered a reasonable interest . This includes, for example, respect for privacy. The Dutch Supreme Court ruled in 1997 that privacy is particularly violated when a publication places a person portrayed in “a public sphere of eroticism and freedom of opinion“. At issue in that case was an advertisement about nude recreation in which someone was portrayed naked unsolicited. Of course, compared to today’s deep-fake porn videos, that is not a big deal. We may assume that this premise certainly applies to such videos.

The privacy interest will still have to be weighed against the right to freedom of expression. Given the serious intrusion of deepfake porn on the privacy of the victims, there is little reason to doubt that the balancing of interests would not turn out in their favour. The victims of deepfake porn videos therefore have a good basis on portrait rights to oppose publications of deepfake porn videos.

GDPR

Besides portrait rights, the General Data Protection Regulation (GDPR) also provides a basis to stand up against deep-fake pornography, as the image of someone’s face qualifies as personal data. Processing personal data requires a valid processing basis. One such basis is consent. It may be clear that this is lacking in this case. Another processing basis may be the legitimate interest of the controller. In that context, a balancing of interests will also be made.

If there is no valid basis for processing personal data, then under the GDPR the data subject has, among other things, the right to erasure of his data – in this case, to erasure of the footage.

Which party to address?

So there are at least two civil law bases besides criminal law to stand up against deepfake porn, but how can victims use these as a basis to ensure that the publications are actually taken offline?

The first step will be to summon the disclosers of deepfake porn to take the publication offline. If it is not possible to identify the discloser, it is possible to write to the host, the party technically responsible for the website on which the footage is published. Does a company or hosting company not voluntarily comply with the summons? Then the court can be asked to impose a publication ban.

There is also the possibility of suing search engines for references to websites hosting the visual material. Indeed, those references may be unlawful in themselves. Moreover, under the GDPR, references in search results from search engine operators constitute their own processing of personal data. Finally, there is the possibility of engaging the Data Protection Authority. Among other things, it can proceed to warn parties or mediate with parties that do not comply with the GDPR.

European Directive

Incidentally, the European Commission is currently working on a proposal for a Directive to combat violence against women. The proposal also regulates something with regard to deep-fake pornography, and for the situation where it involves material withclear similarities to existing persons. Recital 19 reads:

19) […] The unauthorised production or manipulation, for example by image manipulation, of material giving the impression that another person is engaged in sexual activities should also fall within the definition of this offence, to the extent that the material is subsequently made accessible to a large number of end-users by means of information and communication technology without the consent of the person concerned. This includes so-called ” deepfakes ” – material bearing clear similarities to existing persons, objects, places or other entities or events, depicting sexual activities of another person and which would be falsely perceived by others as authentic or truthful. To effectively protect victims of such behaviour, threats of such behaviour should also be included in the definition.

A victim yourself?

In addition to criminal law, there are therefore other options for taking action against publications of deepfake porn. Have you become a victim of deepfake porn yourself? We are happy to help you fight online publications.

Summertime


In the previous episode, among other things, we got rid of the myth of the seven differences. But how do we determine the borderline between inspiring and copying? In other words: at what point is a work no longer a reproduction in modified form, but a new, different work?


From one original to another

I shall illustrate this with a line on which we place a copyrighted work, say a poem, on the left and a totally different copyrighted work, say a painting, on the right. If we consider reproduction as a movement from left to right on that line, there must be a borderline beyond which we can no longer speak of reproduction.

For that left-hand work, for instance, let’s indeed take some lines of poetry, and for that right-hand work a certain painting. As a poem, for example, may serve the following lines, which immediately evoke a warm, somewhat rural image of June/July in the (American) Deep South. (Or nowadays of September in the Netherlands):

Summertime
And the livin’ is easy
Fish are jumpin’
And the cotton is high

(In reality, of course, this is the opening text of the aria “Summertime”, from George Gershwin’s opera Porgy and Bess, with libretto by Edwin DuBose Heyward and Ira Gershwin. But these four lines could well be considered on their own and -certainly if we didn’t know that much more follows- could absolutely count as a “work” in itself (see also this old blog combined with this even much older one).

Reproduction

Those who have read episode 3 of this series know that on the above line, the “ordinary” reproduction, the copy, is very close to the original work. When it comes to a copy of a digital work, you may even wonder whether there is actually any space between the work and the copy. After all, the ones and zeros of the copy are indistinguishable from the ones and zeros of the reproduction.

Modified form

But even the typescript of said lines of poetry (though often no more than a technical reproduction) already looks different from the manuscript with the same lines, and so we move towards the reproduction in altered form. Along the way, we might encounter the work of someone who set the poem to music (Gershwin! – though of course it could have been the other way round). In the case of another poem, this might have involved a single adjustment, to make the metre fit the melody. Recordings may have been made of the song thus created – records, CDs or MP3 or Vobis files. And maybe a music video was also made to accompany it. Or, someone has gone in a different direction and incorporated the lines into a story, or a chapter from a book. These are all potentially reproductions in a modified form.

If we move even further to the right on the line, there comes a moment when we have to say, “OK: now we have a new, different work, which is no longer a reproduction in a modified form. Just past that line, we will first see that the creator has indeed been inspired by the original work (but also no more than that; otherwise we would not have crossed the line).

Inspiration could, for instance, lead to a scat version of the (song) lyrics (“Da padaaa, Dapadada-badahppaa, Da padawaaah, Dapadoebieboewaaa...”). In itself, there is not much left of the poetic lines in this, apart from the metre I have tried to express. However, would this be sung to the tune of “Summertime”, everyone would immediately recognise that. In that case: there would be a replication of the melody, but not (anymore) of the text.

Inspiration could also lead to an image, incorporating the form elements from the lines of poetry. Let’s present this inspiration to Dall-E. Who has absolutely no problem with that. Type in the poem and you get:

Not bad in itself (apart from the incomprehensible letters that were added). That sunhat, in particular, does a lot, and I assume those flowers are the cotton plant in bloom. The fish on the bottom right looks like a catfish, which is typically a fish from the American South. Nicely done, Dall-E!
Still, it is clear at first glance that someone who does not know the genesis of the image would never make the link to “Summertime” here (even as an inspiration).

Other work

As we move further and further to the right on the line above, any resemblance or even memory of the original increasingly disappears, until eventually the works just really have nothing to do with each other at all. For example, we could place the lines of poetry on one side of the line and the (master) work of Prof Visser and Dall E on the other: “blue horse, purple dog & yellow hippopotamus” (from the (now updated!) theses on AI & IE (see episode 2). There is absolutely no relation anymore.

Where is the boundary?

After all this, the key question is, of course, “When do we cross that border?” Has, in our example, the boundary already been crossed if someone, better than Dall E above, does create a painting that perfectly depicts what is described in the lines of poetry -and nothing else? Ordinarily: yes, I would say. With this text at least, that could be counted on. But I could theoretically also imagine texts where we would still remain on the left side of the border. Not easily, but it is possible [1].

What matters is how much “individuality” of the original work can still be found in the other work. In other words, the determining factor is whether 1) copyright-protected elements from the original work, 2) have been recognisably reproduced in that other work. As long as this is still the case, in principle we are still talking about a reproduction in a modified form.

This is a logical criterion. After all, what makes that a creation is a protected work of literature, science or art? That it is original: its own intellectual creation, arising from free creative choices. Whoever adopts such an original, proprietary, element from someone else’s work into another work remains indebted to (the copyright-owner of) that original work. And that stops, of course, when that original element is not (no longer) recognisable. A very small piece of sampled music incorporated into a mix, which is so short that it is not recognised, does not make the mix a reproduction in a modified form. (Then why sample it, one might ask, but that’s another matter.)

Exception: applied arts

There is one exception to the above “rule”: 3) for works of applied art (i.e. industrial design, clothing, etc.), the overall impression is decisive. Here, in this more functional area. infringement is not deemed to exist even if it is noticable that one or more copyrighted elements have been copied, as long as the two products as a whole still give a clearly different impression, infringement is not deemed to exist.

The AI reproducer

So what does this mean for AI copiers? Well: we have seen that at the current state of the art, it is still quite a task to generate text or images that form a reproduction in modified form of an existing work. I have as yet not succeeded. But I will add that the AI generators I work with – ChatGPT (free version) and Dall-E2 – are not exactly top of the bill . Midjourney, in particular, can do much more than Dall-E.

In addition, of course, much (if not everything) is in the prompt. I am therefore willing to believe that there are already programs that (with sharp instruction) could very well deliver products that would qualify as a reproduction in a modified form of an existing, original work.
In any case, we now know how to test that!

But… this is clearly not the only copying the programs do, is it? What about all those existing works that have been “fed” to them, and now enable them to make so many beautiful things. Isn’t this feeding, that input of those works into the machine, also a multiplication? At the moment, in practice, that question is far more important than the question answered here, whether the machine itself can make a reproduction in modified form. High time to look into that (in the next episode)!


[1] It will depend on the degree of (great) originality and detail in the description in the poem and the degree of perfection of depicting that description.