[Skip To Content]
[Website of the National Cancer Institute's Technology Transfer Center.  Partnering with Industry for Improved Public Health.]
  • Home
  • Collaborative Opportunities
  • Standard Forms and Agreements
  • Technology Transfer Training
  • Resources
  • Intellectual Property

Return to Overview

‹‹ Prev Section    Next Section ››

II. Agreements Not To Disclose: Trade-Secrets and the "CDA"

C. Anatomy of a CDA

A normal CDA always addresses four major points, 26

Depending upon the parties negotiating the agreement, it often, but not always, contains some additional terms. Examples of such provisions may include provisions that specify the law of the agreement (e.g., "Federal law shall govern"), certification provisions (e.g., certification by signer of authority to bind the party), indemnification provisions, and disclaimers of warranties. An attorney should be consulted before any of these provisions are accepted. Although these terms may be common, they do not necessarily have to appear in an agreement to make the agreement valid and binding.

in one form or another. First, it identifies the information. Second, it names the parties. Third, it states how the confidential information itself will be handled. Fourth and finally, it specifies the term. Occasionally, some agreements discuss rights to intellectual property -- both that which exists prior to any disclosure under the agreement, and that which is discovered because of the disclosure, should any arise -- but this is not a legally necessary term.

The information to be disclosed defines scope and reach of the agreement. Consequently, this is the single most important part, and a well crafted CDA will clearly identify the information to be disclosed. Unfortunately, there is a tension between the Provider of the information, who typically wants the definition to be as broad as possible, and the Recipient, who wants it as specific as possible. Also, the Provider will not want the CDA’s description of the information to incorporate the confidential information itself.

Nevertheless, some description should be fashioned that will make clear to the Recipient exactly what the Provider expects the Recipient to keep confidential. Also, as a matter of reasonableness, the agreement should specify those situations where information ostensibly provided under the agreement will not be deemed confidential, such as: (a) information which is or becomes public through no misdeed by Recipient; (b) information which Recipient lawfully receives from a third party, which Recipient already knows, or which Recipient independently creates; and (c) information which must be disclosed by force of law.

Next, identifying the parties is simple, yet surprisingly often it is botched by making the individual receiving the information sign as the party, rather than the individual’s employer. One reason this is a mistake is the question of agency -- Providers have essentially no protection if they ask individuals to sign agreements on behalf of their Recipient-employers, unless the individuals’ authority to do so is apparent. Even if agency is not an issue, another problem lies in the hidden trap which caught Gillian Niher when she signed Neurion’s CDA in her personal capacity: She breached her CDA merely by telling her Scientific Director and Lab Chief about Neurion’s information -- not to mention by telling John Rogers at Smallville -- and any measures in breach specified in the CDA could be invoked against her.

How the confidential information will be handled by the parties is usually where the meatiest negotiations occur, because the possibilities are virtually endless. For example, though parties typically agree that written materials claimed to be confidential will be marked as such, how will oral disclosures be treated? What measures will be taken to control who at Recipient’s lab will have access to documents? When the agreement ends, what will be done with the documents, and for how long will Provider’s rights survive? If the Recipient wants to publish, what steps will Recipient have to take to ensure the publication does not contain Provider’s confidential information? What will Provider’s rights be if Recipient is ordered by a court to disclose the confidential information? Each of these issues could be negotiated, within the policies of the parties.

Finally, the agreement should have a clear, specified ending point. Some Providers ask for (and receive) promises to keep information confidential indefinitely. However, as Ben Franklin once wrote in Poor Richard’s Almanac, "three can keep a secret, if two are dead" -- in other words, the more who know a secret, the shorter its secret status will live. In addition, the dizzying pace at which biomedical technology is advancing strongly implies that the commercial value to a piece of confidential information depreciates rapidly, even if competitors never learn the secret. Consequently, a reasonable term to keep a secret should reflect the true life of the

secret, little more. This is particularly important in the academic world, where the act of dissemination is the source of value for information. The NIH policy is that it will keep information given to it confidential for three years, which can be extended for an additional two years upon request -- subject, of course, to the limitations imposed by the Freedom of Information Act. Even for non-Government parties, only in the most unusual circumstances is it even meaningful to promise to maintain a secret for more than five years.

Intellectual property is only occasionally a true issue. Most parties appreciate the unlikelihood that the Recipient will invent something immediately and directly upon seeing the Provider’s confidential information. Others, comfortable of the strength of their background patent position, do not concern themselves with what might happen if someone improves on the technology. In both of these cases, the agreement will state at most that ownership of patentable discoveries will be governed by patent law, and no licenses are promised.

Still, some Providers (usually small companies having a single core technology in a competitive market) will insist that they be promised certain rights in anything invented by the Recipient as a direct consequence of learning the confidential information. Companies and universities may, under the circumstances of the moment, decide that the benefit is worth the risk, and agree to such a term. The Government can never do so under a CDA. With the singular exception of a CRADA (discussed below), any term in an agreement that purports to promise rights in future Government inventions -- including even the option to negotiate a license -- utterly lacks authority under the law.

Return to Overview

‹‹ Prev Section    Next Section ››

Page Last Updated: 12-02-2008