This post is part of a special series including contributions to the OBLB Annual Conference 2022 on ‘Personalized Law—Law by Algorithm’, held in Oxford on 16 June 2022. This post comes from Wolfgang Ernst, who participated on the panel on ‘Personalized Law’.
As a contract law historian, I shall look into personalized contracts from a historical perspective.
Archaic contract making is of a personalized nature. The negotiating parties read signals they hope will disclose the willingness to pay, and they factor all kinds of personal information into their negotiating strategies. Anyone who has found himself in the clutches of a Mediterranean carpet seller knows what I am talking about. The inherent personalization is often a tool to take advantage of the other side.
Standardized, or de-personalized, contracts can be seen in legal contract systems of antiquity already. The most prominent use of standard terms was and continues to be contracts made by auction. Bids need to be made based on all other contract terms fixed to ensure their easy comparability. Different bids can only be ranked if all other parameters of the contract are strictly the same for all bidders. For the same reason, we see invitations to tender for public contracts going out with a fixed set of unchangeable contract terms. Tenders are often coming in under seal. Such contracts are de-personalized by design. Stock market transactions come to mind as well.
The victorious rise of standard contracts was driven by other motives: efficiency and the wish to secure a list of slightly more favourable contract terms which may prove helpful in future disputes about the performance and execution of the contract. A tendency to abuse called for countermeasures which gradually came from judiciaries and legislators. A judicial review of standard terms is now firmly established everywhere.
This made for the distinction between standard terms, unilaterally provided by one party, and contracts individually negotiated (and thus personalized), in whole or in part. As far as contracts are individually negotiated, they escape the judicial review which oversees standard terms. The negotiated (non-standard) contract carries a legitimacy that the de-personalized, unilaterally introduced standard terms lack. The judicial review does not pertain to the price itself. It is presumed that freedom of contract and price agreements should be free from judicial intervention. This makes for a difficult distinction because many parameters of standard terms add to the economic value that is the equivalent of the price. The monetary number that is the price, too, is only one element of the overall economic value that a party has to invest.
All along, one could see different degrees of personalization. Shopping in a department store with fixed prices is a setting that appears, at least at first sight, as largely de-personalized, since all customers get the same terms and prices. Other transactions are highly personalized. Shopping for a Jackson Pollock in a 5th Avenue gallery you will meet personnel that has been trained professionally to spot the real billionaire as opposed to the curious tourist that just wants to take a look. It is not so different from the diligent carpet-seller of old. As a rule of thumb, one will find that mass transactions of small to medium size are just not found worth the trouble of personalization. Cheaper computing power may bring more and more contracts into the range of AI-controlled personalization.
In the past, contracts negotiated tended to be personalized, one way or another. The introduction of algorithms that personalize offers changes this pattern. We now see offers tailored to specific customers, but this tailoring is not the product of an interaction, a negotiation, but a unilateral effort, as unilateral as the standard terms introduced by one party. Information may be used that the offeror has not disclosed deliberately to the other side.
The terms of the contract may appear to be individualized, and not drafted for multiple uses. We need to ask whether a contract thus personalized does come with the greater legitimacy of the contract that emerges from a real negotiation. Do we need judicial review of contracts initiated and tailored using personal data not deliberately disclosed to the offeror? Can the price offered that results from the hidden use of personal information escape judicial review? Should courts have and use the power to lift the unwelcome ‘veil of ignorance’ that hides the use of personal information obtained from one knows not where?
Algorithms may bring progress in another way. To ensure the comparability of tenders or auction bids, it was necessary to restrict bidders to one variable only, the price. It has become possible to factor all sorts of parameters into a formula that allows to rationally rank offers that differ, eg, in price, financing, delivery dates and many more elements. Auction processes that allow bids that differ in a multitude of parameters could bring considerable progress, as bidders could try tweaking their bids in this element or that, playing on their individual strengths. There is considerable potential in that kind of processing of highly personalized offers.
Wolfgang Ernst is the Regius Professor of Civil Law at the University of Oxford.