Author: Gloria Yi Qiao, Andrew Miller
When we talk about machine learning’s application in legal agreements, we have to teach the machine what in-house counsel teams typically fight about and how to resolve such fights depending on the leverage and positioning of the parties. So what do counsel most fight about? And how do they structure an agreement in a fair, mutually beneficial way? Here are some thoughts.
1. Term and Auto Renewal
Typically, the vendors or sellers of an item or a piece of software would want to lock in cash flow via long initial terms and auto-renewing subsequent term agreements. They also want to have long termination notice periods, making it difficult for their customers to opt out of the auto-renewals. Customers typically are not thinking six months ahead, so planning to cancel the auto renew is difficult for them.
If you are the customer, it depends. Sometimes if it’s a hot commodity or a product essential to your operations, you may want to have a longer term with auto renewal to reduce business continuity risk and create more certainty with respect to expenses. However, more often than not, customers want flexibility: 1) a shorter term, maybe just one year, 2) if they agree to no auto renewal, as a compromise, it is with a short notice period for cancellation or termination for convenience at any time on reasonable notice with no penalty.
Agreements that automatically renew also create perverse incentives: they make the supplier feel comfortable that the customer is locked in, reducing motivation to try to improve the product or enhance customer experience.
Even if the vendor agrees to short notice periods for cancellation, without tracking systems in place, the customer may miss the opportunity to exit the contract.
Of course, if you have an intelligent system powered with AI, the system will notify you that an auto-renewal is coming unless you give notice so that you have ample time to deal with it.
2. Termination for Convenience
Another piece that folks fight a lot about is who can terminate the contract when, and on what condition.
If you look at a super power, say Apple’s agreements, you won’t be surprised to find very aggressive termination rights, which we lawyers call “unilateral termination rights”. What this means is that if I am Apple, I can choose to walk away from an agreement any time I choose, for any reason or no reason at all.
If that sounds too much, then the next step down is to allow both parties to walk away for any reason, so that we make the termination rights “mutual”. This definitely gives both parties more freedom, but oftentimes more freedom may not be a good thing. A marriage today is a mutually terminable relationship. Now it doesn’t feel so sexy, huh?
That’s why sometimes the parties may choose to have no termination for convenience. That means, either party can only walk away under certain circumstances. For example, for the supplier, the customers can walk away if the supplier fails to deliver the deliverable on time, or with the quality or quantity promised. For the supplier, they can walk away if the customers fail to pay them. Also, there are some universal situations where the parties are allowed to walk away, for example, when the other party declares bankruptcy.
Now that we’ve gotten some of the simpler terms out of the way, let’s discuss some much “hairier” terms.
The first one is indemnification.
Indemnification means one party of the agreement agrees to pay for the other party’s losses under certain conditions, typically where a third party is harmed and brings a claim. For example, if Acme Co. supplier supplies a part, which Manufacturer Co incorporates into its product, and that part then, in turn, fails and injures Manufacturer’s customers, who then sue Manufacturer, Acme Co. should cover Manufacturer for making things right.
Under some circumstances, indemnification may be mutual. For example, in a software licensing agreement, a software developer might indemnify the licensee against patent infringement claims, while the licensee might indemnify the developer against losses resulting from unauthorized use or modification of the software.
Here again, if one party has a lot of leverage or negotiating power, they may require the other party to indemnify them for everything under the sun and yet relieve themselves from any liability for indemnification. In a more balanced situation, however, the parties may agree to indemnify each other for certain risks within their control.
4. Limitation of Liability
Indemnity deals with third-party claims (i.e. claims by persons who are not parties to the contract). But what about claims between the parties(e.g. for non-payment or failure to deliver, or breach of warranty)? A limitation of liability is a limit that the parties impose as the maximum that a party might be liable for under defined circumstances.
Limitation of liability clauses typically define categories of damages: delay, defect, late payment, and then limit those categories by the concept of proximity:
- Did Acme’s delay mean the manufacturer had to buy elsewhere at a higher price (additional cost = direct damage)?
- Did Acme’s delay mean the manufacturer had to shut down the production line, leaving its employees idle (wages during the delay = indirect damages)?
- Did Acme’s delay mean the manufacturer lost sales (lost profits = consequential damages)?
Some agreements will limit indirect damages. Most agreements will exclude consequential damages altogether.
Among the damages that are covered, businesses want to limit the dollar amount. Sometimes the limit will be determined by how much was paid. For example, in a contract where one million dollars is paid for goods supplied, the parties may agree to limit the total damages to one million dollars or less, or the greater of the total dollar amount paid and some pre-agreed dollar figure.
Indemnity damages and damages for breach of intellectual property terms of confidentiality are typically either not limited or are subject to a much higher limit or “cap”. From the customer’s point of view, even though the supplier may have just only earned one million dollars from the contract, the damages they can potentially cause can far exceed that. For example, if the supplier leaks the customer’s confidential information, the damage can be huge. So, if they agree to a cap, it may be a multiple of the total $ paid under the agreement, such as 5x.
A “greater of” formula is useful to cover “ramp up risk”. So, let’s say a supplier breaches the contract in the first month: liability measured against amounts paid would yield little, if not zero, damages. Where this might be a concern, parties will often settle on a formula such as : “the greater of $[x] and the amounts paid or payable in the twelve month period prior to the event giving rise to the claim”.
When the parties transact over physical parts, or sometimes even software, one thing they may care tremendously about is warranty.
For the supplier, they may want to disclaim all liability and say that the merchandise is provided “as-is”.
For the customer, they may demand a time period for the warranty. In the alternative, the warranty can depend on the wear and tear of the item. For example, in the car world, it’s not uncommon to see a warranty term for the greater of 5 years or 50,000 miles.
6. IP Ownership
For a development agreement or partnership agreement where the parties produce intellectual property (“IP”) as a result of the engagement, the parties may also argue about who owes the IP.
For the IP that each party brings to the table, it’s called “background IP”. That’s easy. What’s yours will be yours and what’s mine will be mine.
The complication arises when the parties produce IP as a result of the agreement (“foreground IP”), oftentimes paid by the customer. The supplier, if the result produced is closely related to its core business, wants to own what’s produced. So the strongest position they can take is to say that the supplier owns all the IP.
However, for the customer, if they paid for the development, why shouldn’t they own the result of it? So the customer will at least claim joint ownership of the IP (some may go for sole ownership, but that’s extremely aggressive and unlikely to be agreed to by the supplier).
If a joint ownership is agreed upon, the parties will also fight for a royalty-free, perpetual license for the portions that they may not own entirely so that they may use the entire end product for other purposes without ever having to get permission or pay the other party. Joint ownership may be dealt with differently under the law in different jurisdictions so joint ownership of IP can be a complex solution, and one that is best avoided.