The dominant model for remote notarial work in Europe is a video call in which the client uploads and downloads their own documents. For a notary, or for the firm that holds the appointment, that means the evidential value of the act still rests on a human looking at a face and an identity document over a compressed video link. No European legislator requires that arrangement, and the reason it persists anyway is more interesting than professional resistance to technology.
The trend moves the meeting, not the identification. Notaries across Europe now handle matters by video call and post, and in several member states video-based certification is the statutory main route. The client appears on camera with an identity document, signs, scans the document and sends it in. The notary receives the file, certifies it and sends it back.
What that produces is the paper process with a camera pointed at it. The identification is a human judgement about image material whose quality neither party controls, the document chain runs through the client's own mailbox and downloads folder, and the finished instrument carries nothing a third party can test afterwards. It records what the notary perceived, and no data about how the person was identified.
That gap is what a qualified electronic identity exists to close. The chip in the identity document is read, the check is recorded, and the identification is cryptographically bound to a certificate that travels with the signature, which under Article 25(2) of the eIDAS Regulation has the equivalent legal effect of a handwritten signature. The difference from the video model is not convenience. It is whether the identification survives in the document or only in the notary's file note.
Less often than practitioners assume. The Swedish Act on Notaries Public (1981:1363) sets out in section 1 what a notary may do: attest signatures, copies, translations and other particulars of a document's content, witness the sealing and opening of storage, report on observations, take declarations of legal or economic significance, and confirm authority to act. It describes the acts. It does not prescribe how the person in front of the notary is to be identified, nor that this happens in the same room, and the accompanying Ordinance (1982:327), which governs appointment, geographic area, deputy notaries and the law degree requirement, is silent on the point too.
That the signature must be made in the notary's presence follows from the nature of the act. But presence and physical co-location are not the same thing in the statutory text, and the familiar list of acceptable identity documents is professional practice rather than legislation. Before concluding that your own jurisdiction differs, read your enabling legislation the same way. The question is how much of what you treat as law is in fact habit.
Partly because no body has a mandate to standardise digital notarial practice: appointments are individual, and what becomes standard becomes so through habit. Mainly because the legislator runs two tracks in opposite directions, and eIDAS expressly permits it. Two provisions in the Regulation's second article do the work, and they are rarely cited in the same breath as Article 25(2). Article 2(3) provides that the Regulation does not affect Union or national law relating to the conclusion and validity of contracts, other legal or procedural obligations relating to form, or sector-specific requirements relating to form, and the 2024 amendment widened that carve-out by adding sector-specific form requirements. Article 2(2) provides that the Regulation does not apply to trust services used exclusively within closed systems resulting from national law.
Read them together. Article 2(3) preserves the form requirement, Article 2(2) exempts the closed national channel, and between them lies exactly the space notarial work occupies. The fragmentation is not an oversight in the framework. It is licensed by it.
The obvious objection is that a public body cannot simply decline a qualified signature, and there is a provision on that. Article 27 obliges public sector bodies requiring an advanced electronic signature for an online service to recognise advanced and qualified signatures conforming to Commission Implementing Decision (EU) 2015/1506, and bars them from demanding a level above qualified for cross-border use. But a register requiring notarial certification in a prescribed national form is not requiring an electronic signature for an online service. Article 2(3) answers first and Article 27 never engages, which is why arguing the point on its merits tends to fail.
Badly, and in two ways that are routinely conflated. Either the form rules refuse electronic form outright, or they mandate digital form through a national channel that will not accept identities from other member states. The second is the more damaging, because it looks like progress.
Germany is the clearest case of the second type. Filings to the commercial register must be submitted electronically in publicly certified form under section 12(1) of the Commercial Code, which expressly permits certification by video communication under section 40a of the Notarial Documentation Act. So the register is not a paper register. But certification may only run through the video system operated by the Federal Chamber of Notaries, an arrangement the Chamber describes as Germany choosing a sovereign design when implementing the Digitalisation Directive, as many member states did.
The most instructive version of the problem is not in any statute. Earlier this year we wound up a German subsidiary and filed the register application and the apostille electronically signed and certified. We were told the form would not be accepted and that both would have to be signed and certified in wet ink, which is what we did. Nothing in the chain was invalid. The signature was qualified, the apostille came from a Swedish notary, and paper was still the rational course, because nobody with a live registration at stake will spend it testing a point of European law.
Read that as a market signal rather than a legal one. It is what fragmentation costs: not a prohibition, but an unwritten expectation that no party will litigate on its own time. A rule you can read is a rule you can plan around. A practice you discover only by filing and being refused is a reason never to file.
Then comes the decision that makes the consequence concrete. By order of 25 February 2026 in case II ZB 13/24, the German Federal Court of Justice held that Austrian online certification is not substantively equivalent to the German procedure, so a qualified electronic signature certified online by an Austrian notary did not satisfy the form requirement and the register court's rejection stood. The German regime requires an electronic identification means at assurance level high, personal identification by the notary against an electronically read photograph, and a state-operated video system. Austrian law falls below this in several respects, including by permitting identification with a physical document that has no electronic function.
Note what that ruling is actually saying. The objection to the Austrian procedure is the objection to the video model generally: identification against a plastic card held up to a camera rather than against a chip that has been read. When a member state takes identification quality seriously it converges on qualified identity, and then locks it inside a national channel.
Sweden is the clearest case of the first type. Lantmäteriet's own guidance on registration of title states that the acquisition document must be signed in writing, that it cannot be signed electronically via the national eID or any other signing solution, and that electronic signatures are not valid for the conveyance of real property under the requirements laid down in law. The same page instructs applicants to sign originals with a blue ink pen. It also records that from 1 July 2026 applicants must identify themselves with a Swedish personal identity number, coordination number or organisation number, to strengthen security in the property market and ensure correct identity data is documented when a property changes hands. Same year, same authority: the identity requirement at intake tightened, the blue ink pen retained at output. The state's own inquiry has already named the answer, since SOU 2024:38 proposes amending the Land Code to permit electronic conveyancing documents at the qualified eIDAS level, and named that same 1 July 2026 for entry into force. The form requirement remains.
The fragmentation is conceded in the directive itself. A recital to Directive (EU) 2019/1151 records that some member states allow company formation and register filings only in person, some allow either route, and in others they can only be done online. The same directive requires member states to make electronic identification and trust services work for cross-border users under Regulation 910/2014, and to report how many cases still required physical presence.
| Jurisdiction and matter type | What the legislator requires | What is still required in practice |
|---|---|---|
| Sweden, conveyance of real property | Strengthened identity requirements for registration of title from 1 July 2026 | Acquisition document signed in writing, electronic signature not valid, originals signed with a blue ink pen |
| Germany, commercial register | Electronic filing in publicly certified form, video certification permitted | Certification only via the Federal Chamber of Notaries system; power of attorney may be required in paper original |
| Germany, cross-border | eIDAS: a qualified signature has the effect of a handwritten one across the Union | Austrian online certification rejected by the Federal Court of Justice, February 2026; a foreign electronic filing and apostille expected in wet ink as unwritten practice |
| Austria | Online certification permitted | Identification allowed with a physical identity document lacking an electronic function, the basis for the German rejection |
| Union level | Directive 2019/1151: company formation must be possible fully online | A recital concedes the spread; member states must report how many cases required physical presence |
One register, three answers, depending on where the identity came from. A domestic filing certified through the national channel is accepted, an online certification from another member state is not, and a foreign electronic chain is expected on paper. The form question therefore cannot be answered in general. It is answered per recipient and per matter type, and the answer changes without anyone announcing it.
Valid yes, accepted not always, and the gap between the two is where the work sits. According to the Hague Conference's Apostille Section, an e-Apostille may not be refused simply because it was issued in electronic form: the same rules apply as for a paper apostille, and it must be accepted by all 130 Contracting Parties. Implementation is another matter. Per the Permanent Bureau's report to CGAP in March 2026, 59 Contracting Parties have implemented one of the two e-APP components and 42 have implemented both, and since 2025 the Conference has been working on a mandate concerning acceptance of e-Apostilles where a Contracting Party's domestic law cannot receive electronic public documents. Recognised at convention level, unsolved.
Two nuances are routinely lost. The apostille itself does not require the client to be present, the underlying attestation does. And in Sweden only a notary public may issue an apostille, which makes the profession the sole route into that system for Swedish documents.
The practical conclusion from all of the above is that you can digitalise the identification without digitalising the output format. Qualified, recorded and verifiable identification is possible even where the finished instrument must be printed, stamped and posted. That is not a compromise. It puts the evidential weight where it does work and accepts the form requirement where it is real.
| Attestation across the counter | Attestation by video call | Attestation with qualified identity | |
|---|---|---|---|
| Basis for identification | Physical document, visual check | Image of document in a video stream | Chip read from the document, recorded check |
| What survives in the instrument | The notary's assertion | The notary's assertion | The assertion plus verifiable identity data |
| Testable by a third party later | No | No | Yes |
| Document chain | Under the notary's control | The client's email and downloads | Controlled storage with an access log |
| Geographic reach | Travel distance | Global | Global |
| Dependent on the recipient's form rules | Low | High | High for output, none for the identification |
Raising the bar at intake reduces intake. That is not a side effect to be talked around, it is the actual trade. A share of clients will not complete a qualified identification and the practice loses them, a cost that arrives immediately and shows up in the diary. The cost on the other side, identity fraud supported by AI-generated material, regulatory penalties and damaged standing, is probabilistic and arrives later.
A practice that has not priced the second side will always rationally refuse the first, which is why arguments about what the law permits so rarely change anything. If you believe the fraud risk is carried by someone else, you are quite right to treat intake friction as a pure loss.
So the relevant question is economic rather than legal. A notary sells one thing: the assertion that a named person appeared and signed. If that assertion turns out to be wrong once, in a matter with real economic substance such as a power of attorney used to dispose of property abroad, it is not the individual document that is questioned but the method, and therefore every act performed by the same method. Where the identification was made over video, there is nothing to produce beyond the file note describing what was seen.
Two things recur in our meetings with law firms. The heaviest friction sits in client onboarding, particularly for one-off matters, which is almost exactly what notarial work consists of. And document exchange outside the firm still runs on email, occasionally on registered post, even at firms that are otherwise well digitalised. Digitalisation stops at the client boundary, which is where evidential value is created.
| Cost of friction at intake | Cost of an identification that does not hold | |
|---|---|---|
| Timing | Immediate | Later, at a single event |
| Measurability | High: abandoned matters | Low before, total after |
| Borne by | The practice | The practice, the client and the appointment |
| Scope | One matter | Every matter using the same method |
| Insurable | Partly | No, standing cannot be underwritten |
First, ask which body will receive the act and what form it requires for that matter type. That determines whether the output must be paper, not whether the identification must be analogue.
Second, ask what you could produce if an attestation were challenged two years later. If the answer is a note recording that identification was presented, that is the answer you have.
Third, ask what a single failed matter would cost the practice, including the appointment. Put a number on it, because without one the friction argument wins every time.
Fourth, ask who carries the risk today, and if the answer is the recipient, check whether the recipient agrees.
For practices that conclude the identification should be qualified even when the output is paper, Trust Circle is built for that arrangement: regulated identities under the anti-money-laundering framework, controlled client access, document storage in Sweden, and qualified electronic signature as one component among them. Because those identities rest on ZealiD's standing as a Qualified Trust Service Provider on the EU Trusted List, the identification survives in the document rather than only in the recollection of the meeting.
If you want to see how this maps onto your own practice, get in touch. Twenty minutes is enough to establish whether it is relevant.
Regulation (EU) No 910/2014 of the European Parliament and of the Council on electronic identification and trust services for electronic transactions in the internal market, Articles 2(2), 2(3), 25(2) and 27, consolidated version. European Union, 2024. https://eur-lex.europa.eu/eli/reg/2014/910/2024-10-18/eng
Handelsgesetzbuch (HGB), Section 12(1). Bundesministerium der Justiz. https://www.gesetze-im-internet.de/hgb/__12.html
Beurkundungsgesetz (BeurkG), Section 40a. Bundesministerium der Justiz. https://www.gesetze-im-internet.de/beurkg/__40a.html
Jordabalk (Land Code) (1970:994), Chapter 4, Section 1. Swedish Parliament. https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/jordabalk-1970994_sfs-1970-994/
European Commission. "What is the legislation, eSignature", on Articles 27 and 37 and Commission Implementing Decision (EU) 2015/1506. https://ec.europa.eu/digital-building-blocks/sites/spaces/DIGITAL/pages/467109076/What+is+the+legislation+-+eSignature
Directive (EU) 2019/1151 of the European Parliament and of the Council amending Directive (EU) 2017/1132 as regards the use of digital tools and processes in company law. European Union, 2019. https://eur-lex.europa.eu/eli/dir/2019/1151/oj/eng
Act on Notaries Public (1981:1363), section 1. Swedish Parliament. https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/lag-19811363-om-notarius-publicus_sfs-1981-1363/
Ordinance on Notaries Public (1982:327). Swedish Parliament. https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/forordning-1982327-om-notarius-publicus_sfs-1982-327/
Lantmäteriet. "Registrera ägare, lagfart" (Registration of title). Retrieved 11 August 2026. https://www.lantmateriet.se/sv/fastighet-och-mark/kopa-aga-salja-eller-ge-bort/registrera-agare---lagfart/
Digitala fastighetsköp och förköpsrätt vid fastighetstransaktioner, SOU 2024:38. Government Offices of Sweden, 2024. https://www.regeringen.se/rattsliga-dokument/statens-offentliga-utredningar/2024/05/sou-202438/
Bundesgerichtshof, order of 25 February 2026, II ZB 13/24. https://www.bundesgerichtshof.de/SharedDocs/Entscheidungen/DE/Zivilsenate/II_ZS/2024/II_ZB__13-24.pdf?__blob=publicationFile&v=1
Bundesnotarkammer. "Beglaubigungen und Beurkundungen in ausländischen Online-Verfahren." https://www.bnotk.de/aufgaben-und-taetigkeiten/rundschreiben/details/beglaubigungen-und-beurkundungen-in-auslaendischen-online-verfahren
Hague Conference on Private International Law. "Apostille Section, e-APP." https://www.hcch.net/en/instruments/conventions/specialised-sections/apostille
Hague Conference on Private International Law. "Apostille Convention, Preliminary Document No 11 of March 2026." https://assets.hcch.net/docs/78570bfc-49fe-4683-ac18-8f9f1dc50462.pdf
County Administrative Board of Stockholm. "Notarius publicus." https://www.lansstyrelsen.se/stockholm/samhalle/tillstand-for-att-utova-verksamhet/notarius-publicus.html