Everything above assumes an absent author. In contemporary art the author is not absent: the artist is alive and retains rights over the work that survive the sale and that the owner never acquired. The principal one is the moral right to respect for the integrity of the work. It does not belong to whoever pays, it does not transfer with the painting, and it gives a third party a voice in a claim file where nobody expected one. It is the only module in this certification where a stakeholder can object to what insurer and insured decided together.
The first consequence bears on restoration, and it reverses a reflex. The insured must mitigate, therefore have the work repaired; but a restoration that alters the work can infringe its integrity, and the artist may object. The route the insurer prefers, the cheapest one, may therefore be legally closed, and the refusal comes neither from the insured nor from the contract. One must then either find a protocol the author accepts, or treat the file as a loss that restoration cannot reduce, which entirely changes the economics of settlement.
The second consequence is the most interesting and the least anticipated by wordings. The artist may offer to remake or rework the piece themselves. Does the loss then become nil? The remade object is not the damaged one; is it the same insured property, and does the agreed value attach to it? A clause providing for repair or replacement was written for interchangeable goods, and here it meets an object only one person in the world can re-create. Depending on the answer, the insurer owes a sum, an object, or nothing, and the contract almost never decides.
The third consequence is a loss with no physical cause, and it extends what was seen about de-attribution. An artist may disown the work after a restoration they judge unacceptable, and that repudiation destroys the value without any object having been touched since the repair. The mechanism is the same as de-attribution, value resting on a recognition that is withdrawn, but the cause differs: here it is deliberate and comes from an identified person reacting to what has just been done to their work. The policy is no better equipped than in the other case, and the risk is triggered by the preceding loss.
Valuation raises a problem of its own, which is not that of objects with an observable market. A living artist's market is narrow and moves fast, up and down, to the rhythm of a career, an institutional exhibition or a change of gallery. Agreed value therefore ages faster here than elsewhere, for a reason that is not observability but volatility. The revision rhythm suited to an old master is too slow here, and the gap widens in both directions with nothing visible happening.
Editions and multiples must be added, which cover a large share of contemporary production. Destroying one copy of an edition of eight is not destroying a unique object, and it produces an effect the line meets nowhere else: scarcity can raise the value of the surviving copies. If the insured owns others, indemnifying the total loss of one copy comes with an enrichment on their own. That is a genuine problem under the indemnity principle, and it has no standard answer.
In practice a contemporary work's file must carry the artist's position before it is needed. Do they accept a restoration, under what protocol, and by whom. Do they intend to be able to remake the work if destroyed, and on what terms. These questions are asked at acquisition or at loan, calmly, and they are documented. Asking them after damage means questioning someone who is now a stakeholder and affected, about a work they have just seen harmed.
An institution acquires in 2021 a wall work by a living artist, agreed value 480,000 euros. In March, building work damages part of the painted surface. The proposed restorer offers a rework at 55,000 euros. Consulted, the artist objects: he considers the rework would alter the piece, and offers to remake it himself on site, against his production and travel costs, that is 70,000 euros. The institution accepts, the work is remade identically and the artist declares himself satisfied. The insurer then objects that the work existing once more, there is no loss, and that the 70,000 euros are a commission for a new work rather than a claim.
The objection is clever and it fails, though not for the reason the institution instinctively offers. It is clever because it is true that the remade object is not the damaged one, and that a repair-or-replace clause was written for interchangeable goods. It fails because the restoration route, which was the normal route for mitigating the damage, was legally closed: the artist holds a moral right to respect for the integrity of his work, that right never transferred with the sale, and the institution could not override it. The refusal is therefore not a choice by the insured but a binding constraint, and the insured took the only available route, which happens to cost 15,000 euros more than the closed one. On principle, the insurer therefore owes the 70,000 and not zero. Two points remain open that a better-drafted contract would have closed. Does the agreed value attach to the remade work, or must it be re-agreed, given that it is no longer quite the same object? And had the artist instead refused the restoration AND refused to remake, the work would have been economically lost at 480,000 euros with no physical destruction, which is the scenario this file came close to. Both questions are asked at acquisition, calmly, and documented in three lines.
- 01The moral right to respect for the work's integrity is never acquired by the buyer: a third party can object to the restoration insurer and insured chose.
- 02If the artist refuses restoration, the normal route for mitigating damage is closed, and the refusal is a binding constraint rather than a choice by the insured.
- 03A repair-or-replace clause presupposes interchangeable goods: it does not say whether agreed value attaches to a work remade by its author.
- 04An artist's repudiation after restoration destroys value with no physical cause, like a de-attribution, but triggered by the preceding loss.
- 05On an edition, indemnifying the loss of one copy can enrich the insured on those they keep, which the indemnity principle does not accommodate easily.