A common question is whether NLP development qualifies as R&D for the federal R&D tax credit under Section 41. The short answer is that NLP development may constitute qualified research when it involves a technical uncertainty and a process of experimentation. This page explains the framework in general terms. It is educational and is not individualized advice. For the foundational framework, see our page on qualified research.
When This Work May Warrant Review
This work may warrant review when it involves a genuine technical uncertainty and a process of experimentation. Under the four-part test, the work must be for a permitted purpose, be technological in nature, be intended to eliminate uncertainty, and be conducted through a process of experimentation.
Common scenarios that may warrant review include:
- New development — evaluating alternative approaches to resolve uncertainty about whether a new design can achieve the required performance.
- New process — testing alternative approaches to resolve uncertainty about whether a new process can achieve the required performance.
- New application — evaluating alternative approaches to resolve uncertainty about whether an existing approach can perform in a new application.
- New performance target — testing alternative approaches to resolve uncertainty about whether a new performance target can be achieved.
Routine Work vs. Development
A central distinction is between routine work and development:
- Routine work — performing established tasks using known methods and known approaches. There is no technical uncertainty. This is routine work, not research.
- Development — developing new approaches where there is a technical uncertainty about whether the approach can achieve the required performance, and evaluating alternatives to resolve that uncertainty. This may warrant review.
Internal-Use Software Considerations
Where the software is developed primarily for the taxpayer's internal use, the internal-use software rules under Section 41(d)(4)(E) and Treasury Regulation §1.41-4(c)(6) may apply. Software developed for sale, lease, or licensing, or software that enables a non-software business component, may warrant review under the excepted-software rules. The analysis is fact-specific.
Hypothetical Example
Consider a company that is developing a new approach and is uncertain whether any available method can achieve the required performance. The company evaluates alternative approaches, tests each, and systematically varies the approach to resolve the uncertainty. This may warrant review as qualified research.
By contrast, if the same company performs routine work using established methods, that is routine work, not research.
This example is illustrative only and does not state that the activity definitely qualifies.
Documentation That May Help
Records that can help support this work include development records identifying the uncertainty and alternative approaches, test results, and records of how results informed decisions. For more, see our page on R&D tax credit documentation.
Key Takeaway
NLP development may constitute qualified research when it involves a technical uncertainty and a process of experimentation. Implementing standard NLP models generally is not qualified research. Because the distinction is fact-specific, professional review is appropriate before claiming the credit.