1. Waymo LLC v. Uber Technologies (2017–2018)

Facts

• Waymo (Google’s self-driving car division) developed advanced LiDAR-based autonomous driving algorithms and sensor fusion systems.
• Anthony Levandowski, a key engineer, allegedly downloaded over 14,000 confidential files before leaving to form Otto, a self-driving truck startup later acquired by Uber.
• Waymo claimed Uber incorporated stolen trade secrets into its autonomous vehicle program.

Legal Issues

• Whether Uber used Waymo’s self-driving algorithmic trade secrets
• Whether acquisition of Levandowski’s files constituted misappropriation
• Whether Uber had “tainted knowledge” through acquisition

Court Findings / Outcome

• The court found strong evidence of misappropriation risk.
• Uber settled before full trial, paying hundreds of millions in equity and agreeing to restrictions.
• Levandowski separately faced criminal charges.

Key Principle

Even if algorithms are not literally copied line-by-line, possession of confidential technical models, training data, or system architecture can constitute trade secret theft.

2. Google LLC v. Uber / Otto (Levandowski Litigation Context, 2016–2018)

This overlaps with the Waymo case but is important separately for algorithm disclosure principles.

Facts

• Levandowski allegedly downloaded proprietary files containing:
  • LiDAR calibration methods
  • Object detection algorithms
  • Sensor fusion logic for autonomous navigation
• These systems were central to Google’s self-driving stack.

Legal Issues

• Whether engineering knowledge embodied in files counts as trade secrets
• Whether “inevitable use” in a competing system matters

Court Reasoning

• Court emphasized that trade secrets include:
  • Code
  • Design documents
  • Algorithmic methods
  • Data structures used for ML systems
• Even partial memory-based use can qualify if it gives competitive advantage.

Key Principle

Trade secret protection extends beyond code—it includes algorithmic design choices and system architecture, even if reimplemented differently.

3. DuPont de Nemours v. Kolon Industries (2011–2015)

Facts

• DuPont developed highly confidential Kevlar production process algorithms and chemical process controls.
• A former employee working with Kolon allegedly disclosed:
  • Production optimization methods
  • Process flow models
  • Efficiency-enhancing techniques (essentially “industrial algorithms”)

Legal Issues

• Whether process optimization methods qualify as trade secrets
• Whether foreign competitor misuse could be enjoined

Court Outcome

• Jury found Kolon liable for trade secret misappropriation
• Damages awarded exceeded $900 million initially (later reduced in settlement context)
• Court issued injunction restricting Kolon’s Kevlar production

Key Principle

Even non-software “algorithms” (like chemical or manufacturing process logic) are protectable if:

• They are secret
• Provide economic value
• Are subject to reasonable protection

4. Appian Corp. v. Pegasystems (2022)

Facts

• Both companies develop low-code workflow automation platforms
• Appian alleged Pegasystems used a “covert contractor” to access Appian’s platform and extract:
  • Process automation logic
  • Rule-based workflow algorithms
  • System design structures

Legal Issues

• Whether platform logic and workflow engines qualify as trade secrets
• Whether indirect access via contractors counts as misappropriation

Court Outcome

• Jury awarded Appian over $2 billion in damages (later reduced on appeal motion stages)
• Court found Pegasystems improperly acquired confidential system design information

Key Principle

Modern courts treat:

“Workflow engines, decision trees, and automation logic” as algorithmic trade secrets if not publicly known.

5. PepsiCo v. Redmond (1995) – Inevitable Disclosure Doctrine

Facts

• A senior executive left PepsiCo for Quaker Oats.
• He had deep knowledge of:
  • Pricing algorithms
  • Distribution optimization models
  • Market strategy systems

Legal Issue

• No direct theft was proven.
• But could his knowledge inevitably lead to disclosure?

Court Decision

• Court issued an injunction preventing him from working in the new role temporarily.

Key Principle

Introduced/strengthened the inevitable disclosure doctrine:

• Even without copying code or documents, courts may restrict employment if:
  • The employee will likely rely on memory of confidential algorithms
  • Competitive harm is unavoidable

6. BladeRoom Group v. Facebook (2018)

Facts

• BladeRoom developed proprietary data center optimization algorithms
• Facebook allegedly used similar design concepts for rapid deployment data centers.

Legal Issues

• Whether system architecture efficiency models qualify as trade secrets
• Whether independent development was credible

Outcome

• Jury initially awarded significant damages (later reduced in post-trial proceedings/settlement discussions)
• Court recognized protectability of:
  • Cooling optimization algorithms
  • Modular deployment methods
  • Energy efficiency designs

Key Principle

Trade secrets can include engineering optimization algorithms, not just software code.

Core Legal Principles Across All Cases

From these cases, courts consistently hold:

1. Algorithms Are Trade Secrets If:

• Not publicly known
• Provide competitive advantage
• Protected by reasonable confidentiality measures

2. “Algorithm” is Broad

Includes:

• Source code
• Machine learning models
• Training datasets
• System architecture
• Business logic rules
• Optimization methods

3. No Need for Exact Copying

Misappropriation can occur through:

• Memory-based replication
• Partial reuse
• Structural imitation

4. Employee Mobility Is Limited by Trade Secret Law

Courts balance:

• Right to work
  vs
• Protection of confidential algorithms

Final Insight

Modern trade secret law treats algorithms as economic assets, not just code. The legal risk is highest when:

• Engineers move between competitors
• AI models or ranking systems are involved
• Internal design documents are accessed or transferred