Project summary

There are two legal questions I keep encountering. First, if I use AI to help design or build something, does that still count as my own invention, in a legal sense? Second, research that I do on my own time, using my own equipment, separate from my paid job. Does that stay mine?

AI has changed something important here. In the past, a person needed a funded lab, a research title, or years of experience inside an industry, just to begin serious R&D. Now, a single person can actually complete a real technical problem from start to end, alone, using AI as a tool. This is a new kind of situation, and the law has not caught up with what "inventor" or "creator" means in this context.

Right now, most people in this position simply guess. Some claim too much credit for work that AI actually did. Others stay quiet about work that may legally belong to them, because they are not sure. I want to research the actual legal reasoning behind both questions, mainly under US law, since most of the relevant case law and patent guidance is developing there. I want to turn this research into something a normal person can actually use, not just legal theory, but something closer to a step-by-step guide.

I became interested in this because I do independent AI research outside of my job, and I realized I did not actually know the legal basis for why that work belongs to me. I would rather understand this properly than continue guessing, and I believe many other people are in a similar situation without realizing it.

What are this project's goals? How will you achieve them?

Goal 1. Research and confirm the actual legal test for both questions above. For AI-assisted invention, the key question is where the line lies between simply describing a desired result and actually defining a specific method or decision logic. Patent law has always required a human to provide the second one, not just the first. For off-duty research, the key question is what actually determines whether the work falls outside a person's job duties. This includes whether personal resources were used, and whether the person made real technical judgments, rather than simply approving something someone else created.

Goal 2. Turn this legal research into a short and practical guide, not only theory. The output will be organized as follows.

1. Core principles to understand before using AI in creative or inventive work

2. A self-check to understand how exposed or protected a person's current work already is

3. Practical responses for specific situations, for example, when an employer asks about the work, when someone questions whether the work truly belongs to the person, or when the person wants to file or publish something

4. Common misunderstandings on this topic, corrected

5. A one-page summary checklist

This guide is intended mainly for people who are not affiliated with any lab or company, since they have the least access to legal advice, and the most to lose if they guess incorrectly.

Goal 3. Have the draft reviewed by an IP lawyer before publication, since I am not a lawyer myself.

Goal 4. Publish the guide openly, in the same way as my other research, so that it is citable and free to use for anyone, whether an individual or an institution.

To achieve these goals, I will continue the research and drafting that I have already started. During this funding period, I will study the relevant case law in more depth, arrange for outside legal review, revise the draft based on that feedback, and then publish it.

How will this funding be used?

Minimum amount needed to begin this project: $15,000 (approximately 2.5 months, full time)

- Living cost while working on this project full time: $13,000

- AI research tools: $250

- Contingency: $1,750

At this level, legal review depends on finding a lawyer willing to volunteer their time. If I cannot find one within the funding period, I will publish the guide clearly marked as not yet reviewed by a lawyer, rather than hiding this fact or delaying publication indefinitely. The scope will remain limited to US law only.

Target amount: $25,000 (approximately 4 months, full time)

- Living cost while working on this project full time: $20,000

- Paid review by an IP lawyer, in case no volunteer is available: $2,500

- AI research tools: $400

- Contingency: $2,100

This amount is enough to complete the guide properly, arrange for a paid legal review instead of relying on chance, and publish without rushing the work.

Ideal amount: $40,000 (approximately 6 months, full time)

- Living cost while working on this project full time: $30,000

- Paid legal review, possibly including a second reviewer for cross-check: $4,000

- A short paid consultation on how this reasoning changes outside the US, for example in Canada or the EU, to include as a brief comparative note: $2,000

- AI research tools: $600

- A plain-language editing pass and basic layout for the one-page checklist: $500

- Contingency: $2,900

At this funding level, I would be able to add the non-US comparative section, which is the part most likely to be removed at lower funding levels, and give the one-page checklist an actual visual design instead of leaving it as plain text.

Who is on your team? What is your track record on similar projects?

I am the only person working on this project. I am an independent researcher based in Toronto, Canada. I have been researching AI on my own time, separate from any job. I have already published a paper on Zenodo (DOI 10.5281/zenodo.19584393), which discusses why cost and quality both tend to get worse as LLM conversations become longer. None of this research was done as part of any job; all of it was self-directed. My background includes a degree in computer science and Buddhist studies, several years of experience as a backend Java developer in Korea, and IT and planning work at a university. This will be my first time conducting formal legal research. However, reading regulations carefully and identifying the actual rule, rather than the rule most people assume, is something I have already applied successfully in real situations before, not only as a theoretical exercise.

What are the most likely causes and outcomes if this project fails?

The review failure mode

• Cause: I am unable to find an IP lawyer willing to review the draft, or I cannot afford to hire one.

• Effect: I publish the guide without independent legal review, or the project is delayed indefinitely while waiting for one.

• Solution: I have included a budget specifically for a paid review, rather than depending only on finding a volunteer (see the funding breakdown above). If I am still unable to arrange a review in time, I will publish the guide clearly labeled as not yet reviewed by a lawyer, rather than hiding this fact or delaying the project without limit.

The jurisdiction failure mode

• Cause: the legal reasoning may turn out to be more contested, and more dependent on individual facts, than I currently expect, particularly outside the US, since both patent law and employment law vary significantly by country.

• Effect: a guide presented as one simple answer could be misleading for readers outside the US.

• Solution: I will state the US-only scope of the guide clearly at the beginning. At higher funding levels, I will add a short comparative note (see the funding breakdown above), rather than assuming without evidence that the same reasoning applies elsewhere.

The time failure mode

• Cause: I run out of funded time before completing the guide.

• Effect: nothing gets published, even though meaningful progress has been made.

• Solution: I will publish the material as a working draft, clearly labeled as unfinished, rather than publishing nothing at all.

Overall, a guide that is confidently incorrect about legal risk would be worse than no guide at all. This is why each of these solutions focuses on being honest about what has and has not yet been verified, rather than trying to appear complete before the work is actually finished.

How much money have you raised in the last 12 months, and from where?

I have not received any funding from a grant or an institution in the last 12 months. My income during this period has come from a physical labor job at a university. This project has been funded entirely by myself, using my own time, up to this point.