An NDA protects confidential information before uncontrolled disclosure; it does not replace IP protection
Before a research team files a patent application, or before it wants to talk to a potential industrial partner, investor, or contractor, it needs a way to share information without losing control over how that information is used. That is what an NDA is for.
This is because an NDA and patent protection operate on different principles.
An NDA (Non-Disclosure Agreement) obliges the party receiving information to keep it confidential and not use it for any purpose other than the one agreed, for a defined period.
Patent protection is a publicly registered exclusive right to use an invention. Obtaining it requires the invention to be new on the priority date, meaning that it must not form part of the state of the art. Confidential disclosure under an NDA is generally not public disclosure, but a breach of the NDA or an overly broad disclosure can, in practice, undermine the possibility of effective protection.
An NDA does not replace patent protection or the protection of know-how as a trade secret. It also does not extend the time available to file a patent application. It creates a contractual obligation that limits who may learn and use the information before it is formally filed for protection, regulated in a substantive agreement, or deliberately made public.
Signing an NDA does not mean the information is safe
In many research teams, an NDA is treated as a formality to complete before any conversation with an outside party can take place.
- Did the other party sign the NDA?
- Was the document approved by the legal department?
- Can the conversation now go ahead?
In commercialization, the effectiveness of an NDA is judged differently.
- Is the scope of protected information clearly defined, or too broad to enforce?
- Does the confidentiality period run long enough to protect the information until a patent is filed or the substantive agreement is signed?
- Does the NDA cover derivative use of the information such as analysis, notes, improvements, not just documents handed over directly?
- Does the agreement define the permitted purpose for using the information and prohibit use outside that purpose?
- Does it include standard exclusions, such as information already public, already known to the recipient, independently developed, or lawfully received from a third party?
- Are the jurisdiction and enforcement mechanism realistic, given where the other party is based?
- Does the other party have real motivation to honour the obligation, or is the agreement just a formality?
- What happens if a breach occurs? can the resulting harm be demonstrated and pursued?
The scope and form of an NDA depend on the situation it applies to. A handful of variants recur in practice.
Common types of NDA
| Unilateral NDA | Only one party, for example, the university, discloses confidential information, and the other commits to protecting it. A common form at an early stage of talks with a potential licensee or investor. |
| Mutual NDA | Both parties disclose confidential information to each other, for example in a joint R&D project where each side contributes its own know-how. |
| Multilateral NDA | Three or more parties, for example, a university, an industrial partner, and an investor, negotiating a deal structure together, covered by a single confidentiality agreement. |
| Limited scope NDA | Covers only information disclosed for a specific, narrowly defined purpose, such as assessing a technology for one application, rather than the team’s activity as a whole. |
|
Confidentiality clause within a broader agreement |
Instead of a separate document, the confidentiality obligation forms part of a wider agreement such as a license, a research collaboration, or an investment agreement. |
The choice of NDA depends on who is disclosing information to whom, for how long, and for what purpose. An agreement drawn too broadly can be hard to enforce, and one drawn too narrowly fails to protect what actually matters; both hinder commercialization equally.
What determines whether an NDA is effective?
Five conditions usually need to be met.
1. A precise definition of confidential information
The agreement must clearly identify what is protected, ideally by referring to specific documents, data, or categories of knowledge, not general language such as “any information disclosed during discussions”.
2. An adequate duration
The confidentiality obligation should run long enough to cover the time needed to file a patent, close the substantive agreement, or complete the technology assessment. For know-how with continuing economic value, the protection period may need to be longer than several years, provided it is reasonable and enforceable.
3. Disclosure limited to what is necessary
The research team should disclose only as much as is needed at a given stage of discussion, holding back the core elements of the know-how until trust in the partner is confirmed by concrete steps, not just a signature.
4. A realistic path to enforcement
An NDA only matters if a breach can be detected, proven, and pursued in practice, which depends on jurisdiction, the team’s legal resources, and whether the resulting harm can be demonstrated at all.
5. Permitted purpose, exclusions, and no implied license
The agreement should clearly state the purpose for which the recipient may use the information, which information is excluded from confidentiality, and that disclosure itself does not grant a license to use the technology, copy the results, or develop them further outside the agreed purpose.
An NDA is not a substitute for caution. It is a contractual obligation against a specific party, not an exclusive right enforceable against everyone. Even a well-drafted agreement cannot undo information once it has been disclosed, and it cannot guarantee protection if the other party decides to breach it.
Effective use of an NDA does not mean having a signed document before every conversation. It means deciding deliberately what to disclose, to whom, for what purpose, and for how long, and treating the agreement as one of several tools for limiting risk, alongside controlling how much is revealed, securing IP protection where appropriate, and building trust gradually as discussions progress.
