As an inventor you often discover commercially important inventions while developing new products or services. It is now up to you to decide how to protect these inventions. You have two options: file a patent or keep the invention a trade secret. Each option has its unique requirements and benefits. As a patent attorney helping startups, I see these issues arise in many startups with artificial intelligence/deep learning technologies.

interplay of patents and trade secrets for startups

Patents and trade secrets are both important forms of intellectual property (IP) protection for startups (and particularly for AI startups), and they play different but complementary roles in protecting a startup’s technology and competitive advantage.

Patents are a form of IP that give the patent holder the exclusive right to prevent others from making, using, selling, and importing an invention for a certain period of time. Patents can be a powerful tool for AI startups, as they can be used to protect the startup’s core technology and prevent competitors from using the same technology. However, the process of obtaining a patent can be costly and time-consuming, and there is no guarantee that a patent will be granted.

Trade secrets, on the other hand, are a form of IP that protect confidential information that gives a business a competitive advantage. Trade secrets can include information such as algorithms, source code, customer lists, and other confidential business information. Trade secrets can be a cost-effective and efficient way for AI startups to protect their technology, as they do not require the same level of public disclosure as patents. However, trade secrets can be more difficult to enforce than patents, as the startup must take steps to maintain the secrecy of the information.

AI startups can use a combination of patents and trade secrets to protect their technology and competitive advantage. For example, an AI startup may use patents to protect the core technology, while using trade secrets to protect the confidential information that gives the startup a competitive advantage. Additionally, AI startups should consider other forms of IP protection, such as trademarks and copyrights, that may be relevant to their business.

For example, AI startups can use trade secrets to protect various aspects of their data collection, training, neural network architecture, algorithm, and AI output.

  1. Data collection: AI startups can use trade secrets to protect the methods they use to collect and organize their data, as well as the specific data sets they use to train their AI models. This can include information about the sources of the data, the processes used to clean and organize the data, and the specific data sets that are used to train the AI models.
  2. Training: AI startups can use trade secrets to protect the specific methods and techniques used to train their AI models. This can include information about the specific algorithms and techniques used, as well as information about the parameters and settings used during the training process.
  3. Neural network architecture: AI startups can use trade secrets to protect the specific architecture of their neural networks. This can include information about the number and types of layers used, the number of neurons in each layer, and the specific activation functions used.
  4. Algorithm: AI startups can use trade secrets to protect the specific algorithms used in their AI models. This can include information about the specific mathematical formulas and equations used, as well as information about the specific techniques and methods used to optimize the algorithms.
  5. AI output: AI startups can use trade secrets to protect the specific outputs of their AI models. This can include information about the specific results or predictions generated by the AI model, as well as information about the specific processes used to interpret or analyze the output.

It’s worth noting that trade secret laws vary between countries and jurisdiction, thus it’s important to ensure that the specific trade secret laws of the jurisdiction in which the AI startup operates are being followed. Additionally, AI startups should also consider other forms of IP protection, such as patents and copyrights, that may be relevant to their business.

The Most Valuable Asset Many Founders Fail to Protect is Trade Secrets

One of the most valuable assets that many founders fail to protect is their trade secrets. Unlike patents, trade secrets are more difficult to copy or reverse engineer, so it’s vital that a startup protect its trade secrets. Keeping your trade secrets private will help protect your company’s future success.

You can protect your trade secrets by enforcing your NDA. Make sure that your employees are not allowed to use your trade secrets for their own benefit. You should also keep all documents updated. This includes any changes that may happen in the future. You should also set aside some time at every regular company meeting to discuss and update your trade secret policy.

The Uniform Trade Secrets Act was passed in 1979, and it standardized the definition of trade secrets across states. This legislation protects businesses from theft and piracy, as well as other infringements of their trade secrets. It also allows businesses to access federal courts to recover damages in case someone misuses their trade secrets.

Trade secrets include software and other proprietary information that can’t easily be reverse-engineered. Even Google’s search algorithm isn’t patented, but it’s still a trade secret because competitors can see data coming in and out of their site. If Google patented their search algorithm, it would have to surrender its competitive advantage.

It’s important for startups to protect their trade secrets if they want to retain their success. Failure to protect this vital intellectual property can lead to a lack of growth and traction. In addition, failure to protect trade secrets can cause loss of market share and future revenue streams.

The value of the information in a trade secret determines its protection. The higher the value of the information, the better chance of protecting it. Other factors that determine the level of protection for trade secrets include the time and money spent developing the information. If the information is easy to duplicate, the value will be lower. In addition to physical measures, firms also need to put in place contractual measures to ensure the safety of their trade secrets.

In addition to protecting your trade secrets, it’s also crucial to consider patent litigation. Many founders fail to protect their intellectual property, and they risk a loss of their company’s reputation. A patent can protect your algorithm, design of your hardware, a piece of software, a business model, a game, and more.

While patents can protect your technology, trade secrets are your business’s most valuable asset, and it’s important to protect it. Trade secrets can be anything that your company uses for your business. That means everything from your customer list to your product formula. This information can be highly confidential. By limiting access to this information, you can protect your business against competition and poaching customers.

What is an NDA?

An NDA It is a legal contract that protects trade secrets or confidential information. It outlines how you will share confidential information. NDAs are sometimes called confidentiality agreements.

According to the World Intellectual Property Organization, confidential information or Trade secrets are intellectual property (IP) rights on confidential information which may be sold or licensed. In general, to qualify as a trade secret, the information must be:

  • commercially valuable because it is secret,
  • be known only to a limited group of persons, and
  • be subject to reasonable steps taken by the rightful holder of the information to keep it secret, including the use of confidentiality agreements for business partners and employees.

In the US, although trade-secret laws can vary from one state to the next, the Uniform Trade Secrets Act is the gold standard for many states, and the Act defines a trade secret in the following terms:

  • Information, including a formula, pattern, compilation, program, device, method, technique, or process,
  • that derives independent economic value, actual or potential, from not being generally known to or readily ascertainable through appropriate means by other persons who might obtain economic value from its disclosure or use; and
  • is the subject of reasonable efforts under the circumstances to maintain its secrecy.

The unauthorized acquisition, use or disclosure of such secret information in a manner contrary to honest commercial practices by others is regarded as an unfair practice and a violation of the trade secret protection. In general, any confidential business information which provides an enterprise a competitive edge and is unknown to others may be protected as a trade secret.

Trade secrets encompass both technical information, such as information concerning manufacturing processes, pharmaceutical test data, designs and drawings of computer programs, and commercial information, such as distribution methods, list of suppliers and clients, and advertising strategies. When you treat confidential information as your secret, restrict access to a limited need-to-know group, and take steps to keep the information a secret, IP rights are created that inure to your company.

A trade secret may also be made up of elements, each of which is in the public domain, but where the combination, which is kept secret, provides a competitive advantage. Trade secrets may protect other examples of information, including financial information, formulas and recipes, and source codes.

The Value of an NDA

IPs are an important part of a company’s success and are often the basis of its business model. If these secrets are not protected, the company can lose valuable business opportunities. An NDA helps protect this valuable IP by stating that the parties to the agreement are obligated to keep the information confidential for a specified period of time. In addition, an NDA can help prevent misuse of the IP by preventing any third parties from obtaining the information.

Many entrepreneurs use NDAs to protect their intellectual property. They can use them with potential investors, clients, advisors, focus groups, manufacturers, and mentors. In addition, NDAs can help protect inventors at the early stages of their innovation. Often, they are used before a patent and can even be combined with one.

An NDA should define IP clearly. This is critical because vague language can lead to ambiguity, which is counterproductive to protecting IP. For example, an NDA should specifically define trademark, copyright, trade secret, and patent. Otherwise, there is a chance that a business partner could accidentally duplicate an idea, leading to liability.

NDAs are an essential part of many corporate transactions. By limiting the amount of information that a company discloses, it grants both parties more trust in each other. It ensures that sensitive information will be protected and that it will not be used for any illegal purpose. It also protects the intellectual property of the company from potential infringement.

Moreover, NDAs can be relatively painless to sign. However, they need to be enforced. If a company breaches an NDA, it can be sued for breach of contract, and the court can award damages. In some cases, it may be possible to settle disputes through mediation services instead of suing.

What information can an NDA Protect?

It is essential to decide what you want the NDA to protect. It can protect information that is recorded and marked as confidential in some form. It may also protect information shared in meetings and presentations.

In the NDA, you can limit the information and ideas you use for a permitted purpose. This could include the evaluation of your idea, or the discussion about a joint venture. You should specify the purpose of sharing confidential information in the NDA As precisely as possible. The permitted use can be extended later. Unfortunately, it is impossible to reduce the restrictions on your ideas and information later.

Consider how long confidentiality should last

It is common to limit confidentiality to three to five years. They will then be able use and disclose your data. Information that is made publicly in any way can be used and disclosed.

You  may specify that some information may be kept secret for life. Examples of such information  include:

  • non-patentable know-how
  • Lists of customers
  • Personal information of the people involved in a project

You might be asked to sign a document by companies and organizations stating that they will not be held responsible for keeping your information or ideas confidential. If this is the case, you should carefully decide whether or not to disclose your ideas to them.

Types of NDAs

NDAs are possible One way Or Mutual. You can use a one-way  NDA If you only disclose information to another party, and in the mutual NDA, both parties agree to protect each other’s confidential information.

If you and the other party to the NDA are not in the same region or state, you must indicate which law governs the agreement and the location or venue of the trial. You may want to designate an exclusive jurisdiction to enforce the NDA in case of unauthorized disclosures made in another country.

Regardless of one way or mutual NDAs, you may want to have different strictness in your NDA form.  The details of the NDA agreement would changed based on risk tiers when it comes to disclosure of business-sensitive information. 

You can use a simpler NDA for Low Risk people such as investors and US vendors, while you can use a restrictive detailed NDA for High Risk parties such as international suppliers or competitors.

Before you meet

Don’t divulge any ideas or information until all parties sign the NDA and return it. Without an NDA, you are taking the chance that other people could use your ideas and information without your permission.

Always read the NDA carefully if asked by another party to sign. In addition, you should ensure that the NDA does not unduly restrict your future activities.

It is essential to ensure that the correct person signs the NDA. Such signatories can be:

  • A director of the recipient company
  • An officer of the company 
  • Someone senior with signature authority to execute the NDA

It is crucial to record the information you give in meetings and presentations. It would be best to ask people to sign a paper copy to prove that they have seen the presentation or technical drawing.

Keep track of the information you share in informal conversations or discussions. Notify where and when it took place.

NDAs  Have Limitations

There are many traps to avoid when you go the trade-secret route. Trade-secret protection is lost forever if the secret becomes public. You need to do everything “reasonable under the circumstance” to safeguard your secrets. It is determined on a case by case basis what qualifies as reasonable.

This could include having your employees sign confidentiality agreements, making sure that they understand what trade secrets should be kept, locking doors to keep information secure, and ensuring that your company has non-disclosure agreements. Trade secrets do not protect you against independent invention or reverse engineering.

Trade-secrets laws do not protect you from competitors who independently develop the invention. However, your competitors can also figure out how your product works, and create a rival product. As long as they don’t have access to your proprietary information, your rights under trade secrets law are not affected. Your competitors could be held liable for trade secret violations AND infringing the patent if the same invention was covered.

In cases where the invention is to be included in a public product, you might consider taking a hybrid approach. You could file patent applications for certain aspects of your technology while keeping other aspects confidential. For example, you might file patent applications for the unique features of a smartphone app’s user interface but keep the software, artificial intelligence (AI) matrix data, and back-end algorithms that make it work secret.

Consider patenting and NDAs together

Since the Supreme Court’s decision in Alice, many commentators have been quick to jump on the software-patents-are-dead bandwagon. While it has become more difficult to patent software-related inventions due to how Alice is being interpreted both by lower courts and the U.S. Patent and Trademark Office, software patents are still very much alive. With software accounting for swaths of the economy, software patents are needed.

Trade secrets offer many advantages over patents. A patent gives you the right to prevent others from using your invention. However, this protection is only valid for 20 years after the patent application was filed. Your invention will no longer be protected by any other person or company. Contrarily, trade secrets are forever, at least theoretically.

Trade secrets are not subject to expiry as long as they remain secret. Trade secret protection is relatively affordable compared to patent protection.

Trade secrets are not subject to a formal filing process or government fees. A specialist will not be required to prepare your application. You also won’t have to pay any additional fees for obtaining a patent which can be a lengthy and expensive process.

While software can be protected as trade secrets, this strategy may be viable for narrow circumstances. Trade secrets are not subject-matter eligible, unlike patents. Trade-secret protection can be granted to any confidential information that could give you a competitive advantage.

Trade secrets would protect any subject matter that could raise questions under Alice, as long as it is not susceptible to reverse engineering. This includes algorithms, software, and business methods. Information that is not patentable before Alice can be trade secrets. This includes sales and distribution methods, consumer profiles, advertising strategies, client lists, or client or supplier lists.  

You should consider filing patent applications and keeping the inventions in those patent applications confidential.   However, this dual identity cannot last forever, as an issued patent or published application will destroy any trade secrets protection for the invention. To use this strategy, file a provisional application first, and at the end of the provisional application, file a non-provisional application. 

This approach adds an extra year to the 18 month publication timeline.  You can also request non-publication of your patent application, but you can only file non-publication requests at the time an application is filed and you can rescind the request at any time. However, you cannot file such a request when patent protection is being sought simultaneously outside the United States.  

In sum, consider filing patents for your trade secrets to give you belt-and-suspender protection, at least for a limited time.  That limited time may be enough to give you the competitive lift you need. We did get more insights on this in our patenting insights article and our story on entrepreneurs’ experiences on going for a patent.