Non-disclosure agreements (NDAs) show up in places most people don’t expect: job interviews, contractor onboarding, startup demos, research collaborations, even conversations that never become a deal. They’re often framed as simple—“Don’t talk about this”—but real NDAs are a bundle of different promises, and some of those promises can be narrower (or shakier) than the signature page makes them feel.
This guide is for ordinary people who want to think clearly about NDAs without sliding into either extreme: “NDAs are ironclad, I can never say anything again” or “NDAs are meaningless, they’re just scare tactics.” In Taming Gravity terms, the goal is to find the science, not the fiction: separate what’s actually written on the page from what’s asserted, what’s disputed, and what’s simply unknown without jurisdiction-specific legal review by a qualified professional.
Start with the “problem scenario”: why NDAs feel absolute
In real life, people run into NDAs under pressure: an employer hands you a stack of onboarding forms, a vendor says “we can’t show you anything unless you sign,” or a source says they can only speak “off the record” with an NDA attached. The social pressure can be stronger than the legal reality.
Two things create the illusion that NDAs are absolute:
- Asymmetry: one side drafts the document; the other side is trying to get hired, get access, or keep a relationship.
- Ambiguity: NDAs often define “Confidential Information” broadly, then add intimidating remedies (injunctions, attorneys’ fees, liquidated damages). Whether those remedies are available, in what situations, and at what scale depends on the wording, the facts, and the governing law—so you shouldn’t assume the most extreme outcome described by the drafter is the one that applies in a real dispute.
If you’re new to Taming Gravity’s approach, the site’s editorial approach and evidence standards are described on Taming Gravity’s manifesto and evidence standards.
What an NDA actually tries to do (and what it doesn’t)
Many NDAs are trying to accomplish a few concrete, practical things:
- Define a bucket of information the disclosing party considers confidential.
- Limit use of that information (for example, “only evaluate a potential partnership”).
- Limit disclosure of that information (for example, “only share with employees who need to know”).
- Set handling rules (security measures, return or destruction upon request).
- Set a time window (how long confidentiality lasts).
What a document labeled “NDA” may not do by itself—or may do only inconsistently, depending on jurisdiction, facts, and drafting:
- Settle IP ownership questions on its label alone. Some documents called “NDA” also contain invention assignment or other IP clauses. Whether those clauses exist in your document—and what practical obligations they create—depends on the text. (This article does not evaluate enforceability of specific IP language without jurisdiction-specific legal authority.)
- Make “public vs. confidential” questions disappear. NDA disputes can turn on timing and proof, and also on whether the NDA treats certain compilations, relationships, or contextual details as confidential even if similar facts exist elsewhere. The contract language and the facts matter.
- Guarantee “gag order” effects across all contexts. Some clauses are drafted broadly, but what happens in a real dispute can be fact- and jurisdiction-dependent. Without local legal authority, you cannot assume a universal rule about what is or isn’t permitted.
Evidence labels for NDAs: separating records, testimony, and inference
Because Taming Gravity emphasizes careful evidence handling, here’s a way to classify what you “know” about an NDA situation:
- Physical evidence / official records: the signed NDA, the version history, emails attaching it, meeting notes, access logs, file watermarks, and any written confidentiality markings.
- Testimony: “They told me this was confidential,” “I was instructed not to share,” “They said it’s okay to discuss generally.” Useful, but less durable than documents.
- Inference: you infer something is confidential because it seems sensitive or proprietary, even if it wasn’t marked or defined.
- Disputed claims: one side asserts something was confidential; the other side says it was already public or independently developed.
- Speculation: assumptions about what a court would do, or what penalties “must” apply, without knowing the controlling law and the contract’s weak points.
In many real disputes, parties fare better when they can point to records rather than relying on memory or assumptions.
Core elements that tend to matter for enforceability
Important: the points below are a practical “what to look for in the text” list, not a universal statement of what any court will enforce. Outcomes turn on jurisdiction, facts, and drafting.
1) How “Confidential Information” is defined
A workable NDA definition usually tries to be broad enough to protect real secrets but specific enough to be meaningful. Common approaches include:
- Marked disclosures: confidential if marked “Confidential” (and oral disclosures confirmed in writing within a set number of days).
- Category-based: lists categories (technical data, customer lists, pricing, designs, research results).
- Catch-all plus examples: “all non-public information relating to…” with examples.
Potential pitfall: if the definition is so broad it effectively covers everything you ever see or learn, it can become hard to apply consistently. Broadness can increase uncertainty and disputes about scope; what that means legally depends on the forum, the facts, and the exact language.
2) The purpose limitation (the “why”)
Many NDAs restrict not just disclosure but use. That matters because, depending on the language, you could be restricting what you do with the information even if nothing is publicly “leaked.” What counts as prohibited “use” is contract- and context-specific.
Tip: look for the words “Purpose,” “Permitted Use,” or “Evaluation.” If the NDA has no purpose limitation, you may be agreeing to broader non-use language than you expect. That’s not proof of illegality or invalidity—but it’s a reason to slow down, ask questions, and (if needed) get jurisdiction-specific advice.
3) The exceptions clause (often the most important paragraph)
Many NDAs try to exclude information that is:
- already public (not due to your breach),
- already known to you before disclosure (and you can prove it),
- independently developed without using the confidential info (and you can document it),
- received from a third party who had the right to disclose it,
- required to be disclosed by law or court order (sometimes with notice obligations).
In disputes, the fight often becomes: can you show an exception applies? This is where contemporaneous notes, timestamps, and clean documentation matter.
4) Duration: how long the obligation lasts
NDAs vary widely by industry and context. Some specify a set term (for example, a few years), while others try to run longer—especially where the disclosing party believes trade secrets are involved. What duration is reasonable or enforceable is jurisdiction- and fact-dependent, so treat any “forever” language as something to evaluate locally rather than a clear, universal rule.
5) Remedies: injunctions, damages, and attorneys’ fees
NDAs often state that the disclosing party may seek an injunction (a court order to stop disclosure) and recover damages. Some include liquidated damages (a pre-set dollar amount per breach). These clauses can raise the stakes. But their real-world impact depends on the governing law, the contract language, and what can actually be proven about breach and harm.
What NDAs typically can’t (or shouldn’t) do
This is the section where internet advice often gets reckless. The careful version is: there are limits on what NDAs can achieve, and some disputes end with clauses being interpreted narrowly (or argued to be invalid) depending on jurisdiction and circumstances. Because we’re not citing jurisdiction-specific authority here, treat this section as questions and red flags to investigate, not as a guarantee.
Public information is not automatically “safe” (and not automatically “confidential”)
If something is legitimately public—published, widely known, or readily ascertainable—that can matter a lot in disputes about “confidential information.” But the analysis can be more nuanced than “public = free to discuss”: some NDAs define confidentiality to include the fact of the relationship, the fact of negotiations, a specific compilation/selection of otherwise public details, or the non-public context in which you learned something. Timing and proof still matter: what exactly was public, when, and what did you learn through the confidential channel?
NDAs shouldn’t be treated as universal gag orders—yet you also shouldn’t assume you’re protected
People often hear “You can always report wrongdoing” or “An NDA can never stop you.” The cautious version is: some legal systems create carve-outs, protections, or limits in certain reporting/investigation contexts, but the specifics vary and the wording matters. If your NDA seems to demand silence about illegal conduct or safety issues, that’s a sign to slow down and get qualified advice for your jurisdiction rather than relying on generic internet rules.
“General skill and knowledge” is a gray zone, not a shield you can count on
A common dispute: after exposure to confidential information, what are you allowed to do with your own expertise? Many NDAs try to draw a line between protected secrets (specific non-public methods, data, designs) and your general skills and experience. Where that line is drawn can be unclear and is often fact-heavy—and disagreements can arise even when someone believes they’re only using “general knowledge.”
Steel-manning both sides: why NDAs exist, and how they get misused
The strongest case for NDAs (the conventional explanation)
- Trade secret protection: companies and researchers want a way to share non-public information to collaborate without losing control of it.
- Efficient trust: an NDA can make early conversations possible without months of due diligence.
- Harm prevention: premature disclosure can damage negotiations, safety reviews, or competitive positioning.
The strongest case for skepticism (the unconventional—but often realistic—explanation)
- Overbreadth as leverage: some NDAs are drafted broadly to intimidate rather than to protect narrowly defined secrets.
- NDA theater: a party may use confidentiality language to make weak claims sound proprietary, or to control narratives.
- Chilling effects: even a questionable NDA can deter speech because the cost of fighting is high.
The evidence-conscious approach is to assume neither “perfectly legitimate” nor “purely abusive” by default. Instead, assess the text, the context, and what can be proven.
A technician-style checklist: what to do before you sign
Use this as a practical, low-drama process.
Step 1: Identify which NDA you’re signing
- Mutual NDA: both sides may disclose confidential info. Often fairer in early discussions.
- One-way NDA: only one side discloses. Common for interviews, vendor demos, and customer evaluations.
Step 2: Find the scope controls
- Is confidential information required to be marked?
- Are oral disclosures handled carefully (confirmed in writing)?
- Is the purpose specific and limited?
Step 3: Audit the exceptions
If the NDA lacks common exceptions (public, prior knowledge, independent development, third-party lawful disclosure), that’s a practical uncertainty: you have fewer clearly stated “exit ramps” inside the document itself, and more may ride on how a dispute is interpreted under applicable law. If the exceptions exist but require strong proof, plan your recordkeeping accordingly.
Step 4: Check duration and return/destruction terms
Make sure you can comply. For example, “destroy all copies” can conflict with automated backups or legal recordkeeping. NDAs sometimes include carve-outs for archival copies retained for compliance—worth requesting if you’re in a regulated environment.
Step 5: Flag any non-NDA clauses hiding in the NDA
Some documents labeled “NDA” also include:
- non-solicitation of employees or customers,
- non-compete-like language,
- assignment of inventions or IP,
- mandatory arbitration, venue selection, or fee-shifting provisions.
These can be consequential. If your document includes them, treat it as more than an NDA.
Common mistakes people make after signing
Mistake 1: Talking in “generalities” that aren’t actually general
People often think changing nouns makes a disclosure safe (“I won’t name the company, I’ll just describe the technique”). But if the technique, dataset, timeline, or internal decision is not public, it may still be treated as confidential under the agreement—especially if it can be re-identified.
Mistake 2: Sharing with friends or online communities
Even if you trust the person, they may not be within the NDA’s “permitted recipients.” In many NDAs, distributing screenshots, excerpts, or paraphrases outside the permitted circle can create breach risk. Whether it is a breach depends on the contract text and facts.
Mistake 3: Assuming “I didn’t sign” means “I’m free”
Sometimes confidentiality duties come from other documents or relationships (for example, employment agreements, onboarding terms, access-control agreements, or policies you accepted as a condition of access). The existence and enforceability of those duties is jurisdiction- and fact-specific; don’t assume the absence of a standalone NDA ends the analysis.
Mistake 4: Not documenting independent development
If you’re building something in the same space as what you saw under NDA, keep clean records of your work. In disputes, documentation can separate independent creation from alleged misuse.
NDAs and extraordinary claims: why this matters at Taming Gravity
NDAs are often invoked in domains where evidence is already contested—advanced aerospace, breakthrough energy, and claimed gravity-related technologies. Here, NDA talk can become a substitute for data: “I can’t show you the evidence because NDA.”
That doesn’t prove deception. It also doesn’t prove the underlying claim. It means the public is being asked to weigh testimony constrained by contract against the absence of records.
It’s also worth naming a conventional, non-mysterious explanation up front: NDA constraints are often standard risk-management practice (privacy, export control, competitive positioning, contractual duties, or classified programs). Sometimes “no public evidence” is just “not legally/contractually releasable,” not an indicator that the underlying story is strong or weak.
A safe “next step” if you’re worried about an NDA
- Find the exact document you signed (final version) and any attachments or incorporated policies.
- Write down the facts: what you disclosed or plan to disclose, to whom, when, and through what medium.
- Map the disclosure to the NDA text: which clause makes you think it’s restricted? Which exception might apply?
- Collect your records: prior art, timestamps, emails, drafts—anything relevant to public/prior/independent-development exceptions.
- Get qualified advice for your jurisdiction if the stakes are real (employment risk, litigation threats, publication decisions, whistleblowing concerns).
If you need to reach the team at Taming Gravity about editorial standards—especially if you’re trying to share a story without breaking obligations—use the site’s contact page and describe what you can share in documented, non-confidential terms.
Conclusion: treat NDAs like claims that require careful parsing
An NDA is not a spell. It’s a set of written obligations with boundaries: definitions, purposes, exceptions, time limits, and remedies. Some are strong, some are poorly drafted, and some collide with laws or public policy in ways that can narrow them. The most practical posture is neither panic nor bravado: read the text, identify what’s actually covered, document what you know independently, and seek jurisdiction-specific help when the consequences matter.
For a site that tries to “find the science, not the fiction,” NDAs are a reminder that missing evidence can have mundane explanations (contractual limits, corporate risk control) without automatically validating extraordinary claims. The disciplined move is to label what you have—records, testimony, inference—and keep your conclusions proportional to the evidence.
Q&A
Are NDAs always enforceable?
No. An NDA is a contract, and enforceability depends on jurisdiction, the exact wording, and the facts. Some NDAs are enforced as written, some are narrowed, and some provisions can be struck if they conflict with law or public policy.
If information is already public, can an NDA still restrict me from talking about it?
Many NDAs include an exception for information that becomes public through no fault of the recipient. But “public” can be a nuanced category in practice: a contract may address the fact of the relationship, a specific compilation, or non-public context you learned through the NDA channel. Real outcomes often turn on what was public when, what you learned from the confidential relationship, and what you can prove.
What NDA clause should I look at first?
After confirming it’s the final signed version, start with (1) the definition of “Confidential Information,” (2) the “Purpose” or permitted-use clause, and (3) the exceptions section. Those three usually determine the day-to-day boundaries.
Can an NDA stop me from reporting wrongdoing?
Do not rely on generic rules here. Some jurisdictions provide protections or carve-outs in certain reporting/investigation contexts, but the specifics vary and the wording matters. If your NDA seems to demand silence about illegal conduct or safety issues, slow down and get qualified, jurisdiction-specific advice.
How do I protect myself if I’m working in the same field after signing an NDA?
Keep clean records showing independent development: dated notes, drafts, version control, and documentation of what you knew before the NDA. If there’s ever a dispute, your ability to show that your work didn’t rely on the confidential disclosure can be crucial.
Leave a Reply