Inventions Agreements – the Death of Innovation

I was talking to a friend once, and he referenced the Inventions Agreement one must sign to work for a company as a form of modern Indentured Servitude. I may not agree entirely, but it is an interesting and common perspective shared by engineers.

A constant challenge faced by software engineering companies is to stay "innovative." Yet it is almost a law in the industry, that as a company grows in size, innovation reduces and maintenance is more common.

This should not come as a surprise, however, to anybody who looks in the right place. If you work in a tech company, go pull out your employment agreement, and look for a clause on Inventions. Read through it and consider what it says. It is highly likely that it says something to the effect of:

Everything you do, create, or build is the company's, not yours, and anything you did prior which you incorporate into the company's work is also the company's.

On the surface, this makes sense. Except for when you get deeper into the fine print, where the argument is that anything you do or think while employed is owned by the employer, even outside of work. There are varying ways of describing this, but they are in favor of the company, not the individual.

This clause is unbalanced and unfair to the engineer. Furthermore, it is presented when the company has the most leverage over an engineer–when onboarding to a new job and they don't want to rock the boat.

The employer has a base of immense power over the employee at this time–they could deny the individual employment. This touches on the most primitive needs of psychology, as defined in the Maslow's Pyramid of needs. So is it any surprise that a significant majority of people just sign it and cross their fingers?

When I have brought this up with other leaders, the typical answer is, "Well, that's not what it means, we wouldn't really do that." Or, "most people are fine with it."

But I would argue that status quo is not a reason, and most engineers are not fine with it, they bristle at it. It is easy for the robber holding a person at gunpoint to ask for all of the person's belongings, but that doesn't make it right.

If this intent is not really to do what it says, then why is it in the contract?

How would it be received if a city, commissioning a notable sculptor to make a new statue for the town square, had a contract that claimed the rights to any other artwork, sculptures or creative output, while the sculptor worked on the city's project? It seems laughable to think somebody would expect that to be the case.

Consider, however, how this might alter the sculptor's behavior. Would they do other work while employed with the city, knowing the city may claim ownership over it, but it isn't really their intent, and just hope for the best?

Would the artist secretly do other work, hiding it from anybody, so that after their agreement is over they can bring it to light?

Would the artist stop being creative, just finishing the one project for the city, and because they have felt so creatively repressed, just give it a moderate effort, instead of their entire soul?

Or would they do other work on the side, and gladly hand it over to the City as the agreement says?

These scenarios, I hope, help paint a picture of what is happening, and it is all of the scenarios except for the last.

If you want to take more control of your IP, consider what it would take to be an IP Punk

The Death of Innovation

Most software engineering companies would like to increase and improve their Innovation. But I contend this simple clause suppresses it.

If you are in a deciding position to push one of these agreements, and by this point, in this article, you feel this is just the screed of an engineer who doesn't understand the bigger business challenges, then I ask you to consider things from a different perspective.

Most engineers who are forced into this sort of servitude often end up having the creative spark dim, thinking, "fine, I will sign it to get hired, but you as a company will get not an ounce more of creative output than what I'm tasked with doing." I've seen it happen many times over, with different developers.

The irony is that what a company thinks will net them more control over their intellectual property is part of the necrotic cancer that drains their Innovation.

These agreements are drafted by people separated from the creative cycle of innovation, so, understandably, they choose the path of least resistance—for them. Yet they never recognize the downstream impact.

One of the first questions I ask an engineer interviewing for a job is what they do as side projects. Those engineers who are willing to admit they have two or three projects are also the most valuable. It shows a mindset of continual learning, reaching, and striving. These are the very people a company wants the most, as they are the ones who can forge through the creative morass and provide new and innovative improvements for the company. Yet they are also the ones who are stifled and go into skunkworks for side efforts.

But consider how different it would be if the company instead encouraged side projects, in a manner that allowed the individual to own their IP?

Can we instead have an agreement that is a little more balanced?

Three Agreement Structures

If you distill the actual claims being asked for an inventions agreement, it boils down to three fundamental positions:

  1. All Company Owned — The company owns it all, with no accounting for exclusions or individually owned inventions. This is by far the worst style, yet one of the more prolific clauses. It works by asserting that while working for the company, they own the creative output that you create.
  2. 1-way Exclusive — An engineer can exclude their IP, but the company owns anything else. Companies who are trying to be better about this often shift to this position. The engineer can exclude prior art, but anything more from that point forward is the companies. Often the space provided to exclude prior art is about 3 lines, squeezed into the bottom of a page— but the better agreements allow for an attached schedule. The challenge with this is it doesn't account for individual IP going forward, only in the past.
  3. Mutually Inclusive — This agreement is the most balanced. The engineer lists prior art which is clearly out of scope, and the company includes the products and scope of work that they are working on. The engineer's work is expressly excluded (to avoid any confusion for Derivative works and licenses granted to the company), and the company's IP is expressly included. The assertion is then that anything worked which is not part of the inclusion is outside the bounds of this agreement. AND, the agreement can be amended as time progresses and new projects are created (from both sides), with a co-signed amendment.

The third option is ideal. However, I have never seen it in actual use. It puts a larger burden of work on the company to have to spell out their IP and keep it updated. The attorneys like the simple clauses there now, but they are not balanced. A change like this comes with risk because you would be forging new clauses. However, the larger risk may come in from the other methods, especially as Open Source Software grows and becomes more dominant in the world (see below for Derivative Works).

A few more challenges

Derivative Works

Unfortunately, the perspective of ownership is not the end of this problem. The reality is that these agreements also look at IP as some clean and discrete product that rolls off an assembly line. It doesn't account for the actual reality that a majority of creative work is derivative from something else that the engineer either finds or has had from past projects.

If the agreement does not consider background IP that either party brings to the table, and how to license that if/when it is included, then it's leaving itself open for challenges. Especially when open-source software is at the heart of something. If you think this isn't a big risk, look into what happened when Cisco included GPL software into their first-generation IP phone.

I should add that I am a contributor and proponent of Open Source Software, but I also know that a company has to protect itself. Being aware of what is used and how helps keep an intellectual property roadmap clean.

Use of Equipment

I have a friend who owns a violin making shop, Peter Prier & Sons Violins in Salt Lake City. In this shop, they have very specialized equipment for creating the structure of a violin, and some of this equipment is expensive. If one of their workers were to use this equipment, even after hours, to work on a private project, then I could understand an Inventions agreement that had a claim on the resulting violin. It is something the violin maker could not do without that equipment.

But this clause has been stretched, distorted, abused and extended in our modern world so much, that people often don't even remember where it came from (old-world things like making violins).

Several years ago, when leaving a company, I was visiting with the CEO who is a great guy. In this conversation, I mentioned some open source software I had created while employed with the company, and that I wanted to release it at large. He suddenly became very serious, and asked a few more questions which led to him asking, "So did you use the work laptop while creating this software?"

He argued that even though I had been meticulous about separating my work efforts from my own because I had used the work-owned laptop as a terminal to get to my private servers in the cloud, that the company had a claim on the resulting software.

I contend that the use of equipment clauses included with an Inventions agreement is out-dated. There should be a separation of common-use equipment, which are part of everyday life, like telephones, network routing, and workstations. Only when it is something that materially affects the company in cost or other denied services, should it be claimable with an inventions clause.

But this expectation gives lawyers migraines. It isn't the status quo. Because there are not any boilerplate agreements that speak to things in this way, and they are standing at the position of power, they just say no, and use the existing agreements.

The other assertion that is even more obscene is when a company then makes the stance that you can only use a workstation for company work, period. Then there is no confusion, right?

Wrong. While this may be the policy, and I have even pressed for this type of policy, it's completely disconnected from how people work in the modern world. A person's workstation or laptop is a ubiquitous tool that is part of their life and is used for both work and play. How to keep that device secure, when it is brought into a company network, is a conversation I can cover in another discussion— we don't have to let bad computer hygiene press for a policy of one device per use case (work or personal).

The situation with that CEO was resolved completely amicably, by the way. I mentioned to him that if he felt the company had an ownership stake in something I created on my own time because I used their laptop as a terminal to connect to my own cloud devices, then by that same logic shouldn't I have a claim over the company's IP, when I did any work from home, through my networking equipment and hardware?

Why is it that the second claim feels so ridiculous, yet to some people, the first claim does not?

What options do we have?

I hope this can spark some conversation and investigation with people. I am not a lawyer, so I hesitate sharing revised contracts where I have pushed to get the items above addressed. But I hope we can work on getting a new software inventions agreement that is tailored to the modern engineering world, including its ability to address derivative works, and is not so draconian and imbalanced.

With a clear understanding that I am not a lawyer, and this is not legal counsel in any way, I present a few paragraphs that I find address these concerns, in hopes that it will spur people who are attorneys to research and help define a more modern and innovation-friendly agreement. This, of course, would be part of a larger agreement.

This text is also incomplete, as it doesn't clarify the differentiation of company equipment that is common and not unique to the creation of the item, and where it doesn't create a material impact upon the employer (i.e. using a work laptop as a terminal to connect to the cloud).

  1. “Invention”, “Inventions” means all processes, tools, discoveries, inventions, works of authorship, documents, trade secrets, models, computer programs, source code, designs, utilities or other intellectual property and all embodiments thereof.
  2. Company Inventions and Intellectual Property. We do not grant You any rights in or to Our intellectual property except such licenses as may be required for You to perform Your obligations hereunder.
  3. "Your Inventions. Your Inventions belong exclusively to You, including to the extent included in the Company’s Inventions.
  4. "Background Technology. Background Technology means all processes, tools, works of authorship, programs, utilities or other intellectual property that existed prior to performance of the Services and are (i) included in, or necessary to, the Invention, and (ii) either owned by You or licensed to You with a right to sublicense.
  5. “Company Invention” is any Invention which is:
    a. conceived, developed, reduced to practice, or created by You
    i. within the scope of this Agreement and any work related to this Agreement;
    ii. on-time applicable to this Agreement and any work related to this Agreement; or
    iii. with the aid, assistance, or use of any of Our property, equipment, facilities, supplies, resources, or intellectual property;
    b. the result of any work, services, or duties performed by You for Us;
    related to the industry or trade of Ours; or
    c. related to the current or demonstrably anticipated business, research, industry, or development of Ours
    To the extent applicable, such Company Invention shall be deemed a “Work made for hire.”
  6. License for other Inventions. If in the course of this Agreement, You incorporate into Company property or create a derivative of an Invention owned by You or in which You have an interest, the Company is granted a nonexclusive, royalty-free, irrevocable, perpetual, worldwide license to make, modify, use and sell such Invention as part of and in connection with the Company property.
  7. Exclusion of Other Inventions. You may identify an Exhibit A, attached hereto, as a list of Inventions which You have conceived, learned, made or first reduced to practice, either alone or jointly with others, prior to this Agreement with the Company and which You desire to exclude from the operation of this Agreement except as is covered in the section “License for other Inventions.” You may amend this exhibit over time, to help with clarity of ownership, and where other Inventions are not considered Company Inventions.
Brandon Gillespie

Brandon Gillespie

Brandon loves Tech and has a breadth of experience including hands-on Implementation and Leadership Roles across many companies including small startups, enterprises, and the USAF.

Read More