First analysis of the European Parliament’s draft proposal to amend Brussels Ia and Rome II with a view to corporate human rights due diligence.

Update 28 October 2020 see also Chris Tomale’s further critical reflection here.

Update 22 October 2020  see for comparative purposes Jan von Hein’s critical comments on the EP draft work for a Regulation on the civil liability for artificial intelligence. There is a clear tendency in the institutions to draft tailor-made regimes for the PIL aspects of whatever hot topic hits them – it is generally neither a wise nor a necessary move.

Update 12 October 2020 Jan von Hein has weighed in on the debate here. Update 9 October 2020 Giesela Ruhl has further review of the Rome II elements here.

Thank you Irene Pietropaoli for alerting me to the European Parliament’s draft proposal for a mandatory human rights due diligence Directive. The official title proposed is a Directive on Corporate Due Diligence and Corporate  Accountability). Parliament also proposes insertions in both Brussels Ia and Rome II. For the related issues see a study I co-authored on the Belgian context, with links to developments in many jurisdictions.

I do not in this post go into all issues and challenges relating to such legislation, focusing instead on a first, preliminary analysis of the conflicts elements of the proposal.

A first issue of note in the newly proposed Directive is the definitional one.  The proposal’s full title as noted uses ‘corporate due diligence and corporate accountability’. However in its substantive provisions it uses ‘duty to respect human rights, the environment and good governance’ and it defines each (but then with the denoter ‘risk’) in Article 3. For human rights risks and for governance risks these definitions link to a non-exhaustive list of international instruments while for the environment no such list is provided.

The proposed Directive points out the existence of sectoral EU due diligence legislation e.g. re timber products and precious metals, and suggests ‘(i)n case of insurmountable incompatibility, the sector-specific legislation shall apply.’ This is an odd way to formulate lex specialis, if alone for the use of the qualifier ‘insurmountable’. One assumes the judge seized will eventually be the arbitrator of insurmountability however from a compliance point of view this is far from ideal.

As for the proposed amendment to Brussels Ia, this would take the form of a forum necessitatis as follows:

Article 26a
Regarding business-related civil claims on human rights violations within the value chain of a company domiciled in the Union or operating in the Union within the scope of Directive xxx/xxxx on Corporate Due Diligence and Corporate Accountability, where no court of a Member State has jurisdiction under this Regulation, the  courts of a Member State may, on an exceptional basis, hear the case if the right to a fair trial or the right to access to justice so requires, in particular: (a) if proceedings cannot reasonably be brought or conducted or would be impossible in a third State with which the dispute is closely related; or (b) if a judgment given on the claim in a third State would not be entitled to recognition and enforcement in the Member State of the court seised under the law of that State and such recognition and enforcement is necessary to ensure that the rights of the claimant are satisfied; and the dispute has a sufficient connection with the Member State of the court seised.

This proposal is a direct copy paste (with only the reference to the newly proposed Directive added) of the European Commission’s proposed forum necessitatis rule (proposed Article 26) at the time Brussels I was amended to Brussels Ia (COM (2010) 748). I discussed the difficulty of such a forum provision eg here (for other related posts use the search string ‘necessitatis’). The application of such a rule also provokes the kinds of difficulty one sees with A33-34 BIa (including the implications of an Anerkennungsprognose).

Coming to the proposed insertion into Rome II, this text reads

Article 6a
Business-related human rights claims
In the context of business-related civil claims for human rights violations within the value chain of an undertaking domiciled in a Member State of the Union or operating in the Union within the scope of Directive xxx/xxxx on Corporate Due Diligence and Corporate Accountability, the law applicable to a non-contractual obligation arising out of the damage sustained shall be the law determined pursuant to Article 4(1), unless the person seeking  compensation for damage chooses to base his or her claim on the law of the country in which the event giving rise to the  damage occurred or on the law of the country in which the parent company has its domicile or, where it does not have a domicile in a Member State, the law of the country where it operates.

I called this a choice between lex locus damni; locus delicti commissi; locus incorporationis; locus activitatis. Many of the associated points of enquiry of such a proposal are currently discussed in Begum v Maran (I should add I have been instructed in that case).

A first obvious issue is that the proposed Article 6a only applies to the human rights violations covered by the newly envisaged Directive. It does not cover the environmental rights. These presumably will continue to be covered by Rome II’s Article 7 for  environmental damage. This will require a delineation between environmental damage that is not also a human rights issue, and those that are both. Neither does the proposed rule apply to the ‘good governance’ elements of the Directive. These presumably will continue to be covered by the general rule of A4 Rome II, with scope for exception per A4(3).

My earlier description of the choice as including ‘locus incorporationis’ is not entirely correct, at least not if the ‘domicile’ criterion is the one of Brussels Ia. A corporation’s domicile is not necessarily that of its state of incorporation and indeed Brussels Ia’s definition of corporate domicile may lead to more than one such domicile. Does the intended rule imply claimant can chose among any of those potential domiciles?

Locus delicti commissi in cases of corporate due diligence (with the alleged impact having taken place abroad) in my view rarely is the same as locus damni, instead referring here to the place where the proper diligence ought to have taken place, such as at the jurisdictional level in CJEU C-147/12 OFAB, and for Rome II Arica Victims. This therefore will often co-incide with the locus incorporationis.

Adding ‘locus activitis’ as I called it or as the proposal does, the law of the country where the parent company operates, clearly will need refining. One presumes the intention is for that law to be one of the Member States (much like the proposed Directive includes in its scope ‘limited liability undertakings governed by the law of a non-Member State and not established in the territory of the Union when they operate in the internal market selling goods or providing services’). Therefore it would be be best to replace ‘country where it operates’ with ‘Member State’ where it operates. However clearly a non-EU domiciled corporation may operate in many Member States, thereby presumably again expanding the list of potential leges causae to pick from. Moreover, the very concept of ‘parent’ company is not defined in the proposal.

In short, the European Parliament with this initiative clearly hopes to gain ground quickly on the debate. As is often the case in such instances, the tent pegs have not yet been quite properly staked.


(Handbook of) European Private International Law, 2nd ed. 2016, Chapter 8, Heading 8.3.

(3rd ed forthcoming February 2021).




Proposition Walhalla. ‘The algorithms of the law must keep pace with new and emerging technologies.’

Update 18 August 2020 the judgment was overturned upon appeal and a breach of Article 8 ECHR found: [2020] EWCA Civ 1058 Bridges, R (On the Application Of) v South Wales Police. reviewed here.

Update 17 January 2020 the European Commission reportedly has a different view and is preparing a proposal to ban this use temporarily.

‘The algorithms of the law must keep pace with new and emerging technologies’ is the opening sentence of Hadon-Cave LJ and Swift J in R v The Chief Constable of South Wales Police and others [2019] EWHC 2341.

The central issue is whether the current legal regime in the United Kingdom is adequate to ensure the appropriate and non-arbitrary use of AFR (automated face recognition) in a free and civilized society. The High Court finds it is. No doubt appeal will follow. I leave the assessment of the findings (discussing in particular Article 8 ECHR: right to respect for one’s private and family life, one’s home and one’s correspondence) of the Court to others. It is the opening sentence which drew my attention as, inevitably, it did others’. It is a sentence upon which one can hinge en entire regulatory /new technologies course. Must the algorithms of the law (whatever these may be) keep pace with technology?  Or rather, guard against the challenges of same?



Draft European ethics guidelines for trustworthy artificial intelligence.

Update 8 April 2019 for the final guidelines following review, see here.

An ethics-related posting seems apprioprate as last before ‘the’ season.

The relevant European expert group seeks feedback on draft ethics guidelines for trustworthy artificial intelligence.

Chapter I deals with ensuring AI’s ethical purpose, by setting out the fundamental rights, principles and values that it should comply with.
From those principles, Chapter II derives guidance on the realisation of Trustworthy AI, tackling both ethical purpose and technical robustness. This is done by listing the requirements for Trustworthy AI and offering an overview of technical and non-technical methods that can be used for its implementation.
Chapter III subsequently operationalises the requirements by providing a concrete but nonexhaustive assessment list for Trustworthy AI. This list is then adapted to specific use cases.

Of particular note at p.12-13 are the implications for the long term use of AI, on which the expert group did not reach consensus. Given that autonomous AI systems in particular have raised popular concern, most of which predicted in the longer term, it is clear that this section could prove particularly sticky as well as interesting.

For me the draft is a neat warm-up for when the group’s co-ordinator, Nathalie Smuha, returns to Leuven in spring to focus on her PhD research with me on the very topic.