Confidentiality Agreements And Algorithmic Patent Filing Strategy.
1. Confidentiality Agreements (NDAs) in Technology and Algorithms
A Confidentiality Agreement (also called a Non-Disclosure Agreement, NDA) is a legal contract used to protect sensitive information. In the context of algorithmic or software inventions, NDAs are crucial for:
Preventing premature disclosure of proprietary algorithms.
Securing trade secrets before filing a patent.
Protecting business strategies, source code, and data sets.
Key Points in NDAs for Algorithms
Definition of Confidential Information: Explicitly include source code, flowcharts, algorithm logic, datasets, and system designs.
Duration: Typically 2–5 years; must cover the period before patent filing.
Obligations of Recipient: Duty to maintain secrecy, restrict use to evaluation purposes only.
Exclusions: Information already in the public domain or independently developed.
2. Algorithmic Patent Filing Strategy
Patenting algorithms can be tricky due to software patentability issues in many jurisdictions. The strategy often involves:
Filing Provisional Patents:
Protect the invention while delaying full patent application (especially useful for algorithms that are evolving rapidly).
Use of NDAs Before Filing:
Prevents “public disclosure,” which could destroy novelty under patent law.
Trade Secret vs. Patent:
Decide if certain algorithm parts should remain trade secrets instead of being patented, especially if:
Reverse-engineering is hard.
Patent enforcement costs outweigh benefits.
Claim Drafting Strategy:
Draft claims around technical effects rather than just abstract mathematical formulas.
Include system-level applications (e.g., data processing, user experience improvement).
3. Case Law Analysis
Here are five detailed cases relevant to confidentiality agreements and algorithmic patent filing:
Case 1: Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470 (1974)
Issue: Trade secrets and patent strategy.
Facts: Kewanee Oil sued Bicron for misappropriation of trade secrets. The question was whether trade secrets that could be patented should still be protected under state law.
Holding: Trade secret law is complementary to patent law. Disclosure under an NDA does not constitute abandonment of trade secret rights.
Significance: Before filing algorithm patents, NDAs can protect confidential information without jeopardizing potential patent claims.
Case 2: Apple Inc. v. Samsung Electronics Co., 678 F.3d 1314 (Fed. Cir. 2012)
Issue: NDA violations and patent infringement.
Facts: Apple and Samsung exchanged confidential technical information under NDAs. Apple claimed Samsung misused trade secrets and patented concepts disclosed during collaborative discussions.
Holding: Courts upheld that misuse of confidential information can lead to injunctions and damages, even if patent applications are later filed by the breaching party.
Significance: Reinforces the importance of NDAs when sharing algorithmic concepts with potential collaborators.
Case 3: Diamond v. Diehr, 450 U.S. 175 (1981)
Issue: Patentability of algorithms.
Facts: Diehr invented a process for curing rubber using a mathematical algorithm. The patent office initially rejected it as an unpatentable mathematical formula.
Holding: The Supreme Court ruled that a process using a mathematical algorithm applied to a practical application is patentable.
Significance: Shows that when filing patents on algorithms, the focus should be on practical applications, not just the abstract formula. NDAs help protect these applications before filing.
Case 4: Trade Secret Theft – Waymo v. Uber, 2017
Issue: Misappropriation of autonomous vehicle algorithms.
Facts: Uber allegedly used confidential files from Waymo to develop self-driving car technology.
Holding: Uber settled for $245 million, acknowledging misappropriation of trade secrets.
Significance: Demonstrates the critical role of NDAs and proper legal controls before sharing algorithmic code and patentable ideas with partners.
Case 5: Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014)
Issue: Software patent eligibility.
Facts: Alice’s patent on a computer-implemented financial scheme was challenged as an abstract idea.
Holding: The Supreme Court ruled that abstract ideas implemented on a computer are not patentable unless they improve technology.
Significance: Guides algorithmic patent filing strategy: claims must focus on technical improvement, not just the abstract algorithm. NDAs protect these innovations while preparing defensible claims.
Other Notable Points from Cases
NDAs are essential to establish trade secret rights (Kewanee, Waymo).
Patent filing strategies must emphasize practical application (Diamond, Alice).
Collaborations and disclosures require strict legal oversight to prevent loss of rights (Apple v. Samsung, Waymo v. Uber).
4. Practical Filing Strategy Combining NDAs and Patents
Pre-Filing Stage:
Sign NDAs before discussing algorithm with investors, collaborators, or contractors.
Document development dates and source code versions.
Provisional Patent:
File early to secure “priority date.”
Keep some modules as trade secrets if disclosure could weaken protection.
Full Patent Application:
Emphasize system-level implementation and technical improvement.
Reference provisional filing to claim priority.
Monitoring Competitors:
NDAs provide evidence in case of misappropriation.
Enforceable in courts, as shown in Apple v. Samsung and Waymo v. Uber.
✅ Conclusion:
NDAs protect your algorithmic ideas before patent filing.
Patent strategy must consider timing, claims, and technical application.
Case law shows that trade secret protection, careful disclosure, and practical implementation claims are essential to defend algorithmic innovations.
RELATED Blog
comments