Balancing Innovation Incentives With Public Accessibility To Algorithms.
I. Introduction
Algorithms are the backbone of modern technology, from search engines to AI systems. Protecting them can:
Incentivize innovation (reward developers, protect investments)
Encourage commercialization
But excessive protection can restrict public access, impeding further innovation, interoperability, and scientific progress. Patent law, copyright law, and trade secret law create overlapping but sometimes conflicting incentives.
Balancing these goals has been central in courts, especially in cases where software, machine learning models, or algorithmic methods are involved.
II. Patent Law and Algorithmic Innovation
1. Alice Corp. v. CLS Bank International (2014)
Facts:
Alice Corp. claimed patents on a computerized scheme for mitigating settlement risk in financial transactions.
Holding:
The Supreme Court held that abstract ideas implemented on a computer are not patentable unless they contain an “inventive concept” beyond the abstract idea itself.
Impact:
Algorithms are often considered abstract ideas.
Mere implementation of an algorithm on a generic computer does not justify patent protection.
Balances innovation incentives (patents are available if the algorithm is truly novel and inventive) with public accessibility (abstract ideas remain free to use).
Key Insight: Encourages publication of fundamental algorithmic ideas while protecting only inventive applications.
2. Gottschalk v. Benson (1972)
Facts:
A patent application claimed a method for converting binary-coded decimal numbers to pure binary using a computer.
Holding:
The Supreme Court denied patentability because the method was a pure algorithm, which is a law of nature or abstract idea.
Impact:
Reinforces that basic algorithms cannot be monopolized.
Protects public access to foundational computational methods.
Courts distinguish between mathematical formulas (unpatentable) and practical applications of algorithms (potentially patentable).
Balance: This ruling maintains broad public access to algorithms while leaving room for patenting applied technological implementations.
3. Diamond v. Diehr (1981)
Facts:
Diehr patented a process for curing synthetic rubber using a mathematical formula implemented in a computer.
Holding:
The Court allowed the patent because the algorithm was applied in a process producing a physical and useful result.
Impact:
Highlights that an algorithm embedded in a technical process can be patentable.
Incentivizes innovation by protecting practical applications of algorithms while keeping pure formulas free for public use.
Balance: Protects inventors’ incentives without unnecessarily restricting public access to the underlying mathematics.
4. In re Bilski (2008)
Facts:
Bilski attempted to patent a method for hedging risks in commodities trading.
Holding:
The Federal Circuit rejected the patent because it was an abstract idea. The Supreme Court later affirmed that abstract methods are not patentable.
Impact:
Reinforces that business methods and abstract algorithms cannot be monopolized.
Encourages public access to algorithmic methods while protecting specific practical applications.
III. Copyright Law and Software Algorithms
5. Computer Associates Int’l, Inc. v. Altai, Inc. (1992)
Facts:
Altai copied portions of Computer Associates’ software.
Holding:
The court developed the “abstraction-filtration-comparison” test to determine what aspects of software code are protected by copyright.
Impact:
Underlying algorithms are not copyrightable, only the expression of code is protected.
Protects innovation incentives for writing software code while keeping algorithms publicly accessible.
Balance: Encourages creative coding while allowing other developers to implement similar algorithms differently.
6. Oracle America, Inc. v. Google, Inc. (2018)
Facts:
Oracle claimed copyright infringement over the use of Java APIs in Android.
Holding:
The Supreme Court ruled that Google’s use of Java APIs was fair use, considering it transformative and necessary for software interoperability.
Impact:
Protects software interoperability—a key public-access concern.
Prevents monopolization of widely-used interfaces that are essential for innovation in other software systems.
Balance: Encourages innovation (developers can build new platforms) while preserving access to widely-used algorithmic interfaces.
IV. Trade Secrets and Public Accessibility
7. Kewanee Oil Co. v. Bicron Corp. (1974)
Facts:
Kewanee attempted to claim patent protection for trade secrets but had previously relied on secret formulae.
Holding:
The Supreme Court recognized that trade secrets are legally protectable even without patenting but cannot prevent independent invention.
Impact:
Protects commercial incentives for keeping algorithms secret.
Does not block public access entirely—once disclosed or independently discovered, the algorithm enters the public domain.
Balance: Encourages private R&D while preserving eventual public access.
8. Ruckelshaus v. Monsanto Co. (1984)
Facts:
Monsanto’s confidential pesticide formula was submitted to the government. Ruckelshaus later disclosed it in a regulatory context.
Holding:
The Supreme Court held that trade secrets must be protected unless public interest justifies disclosure.
Impact:
Illustrates tension between private innovation incentives and public health and accessibility.
Algorithms for critical areas (e.g., AI in healthcare) may eventually need public access for safety and innovation.
V. Emerging AI and Algorithmic Cases
9. Thaler v. USPTO (DAB 2022)
Facts:
Stephen Thaler claimed patent inventorship for an AI-generated invention.
Holding:
The USPTO rejected AI as an inventor, requiring human inventors.
Impact:
Highlights policy balancing: protect human innovation incentives while keeping AI-generated algorithms accessible.
Avoids giving monopolistic rights to entities over AI-generated methods that could otherwise benefit the public.
VI. Key Takeaways and Balancing Principles
Abstract ideas vs. applied algorithms
Gottschalk, Alice, Bilski: Abstract algorithms remain free for public use.
Diamond v. Diehr: Applied methods can be patented.
Software expression vs. functionality
Computer Associates v. Altai, Oracle v. Google: Protect code expression but not underlying logic.
Trade secrets vs. eventual public access
Kewanee, Ruckelshaus: Protect proprietary algorithms temporarily but allow eventual public access for broader societal benefit.
AI-generated work
Thaler: Human inventorship required, maintaining balance between incentives and accessibility.
Overall Principle:
Patents, copyright, and trade secrets incentivize innovation.
Courts carefully limit exclusivity to avoid monopolizing abstract ideas or algorithms essential for further innovation.
Public access is preserved either by:
Rejection of abstract claims
Limiting copyright to expression
Requiring eventual disclosure of trade secrets in regulated areas
RELATED Blog
comments