LYNCH·LLP
Book a consultation →
HOME/INDUSTRIES/AI, SOFTWARE & SAAS
◈ INDUSTRIES / AI, SOFTWARE & SAAS

Software-eligible.
Defensible.

Patents that survive § 101, copyrights for source code, trade secrets for model weights, and trademarks that travel — all coordinated for the way modern software companies actually ship.

// 01 / IP CHALLENGES IN THIS SECTOR

What we see, day to day.

◈ 01 / CHALLENGE
§ 101 subject-matter eligibility
Generic 'do it on a computer' claims don't survive Alice. We draft to the technical-improvement and specific-implementation prongs.
◈ 02 / CHALLENGE
Trade secrets vs. patent disclosure
Model weights, training pipelines, and ranking signals are often better as trade secrets. Knowing which to file and which to lock down matters.
◈ 03 / CHALLENGE
Open-source license compatibility
AGPL, GPL, MIT, Apache — incompatible licenses in your dependency tree can blow up an acquisition.
◈ 04 / CHALLENGE
AI-generated content + training-data risk
Output ownership, training-set licensing, and the rapidly moving copyright landscape around generative models.
// 02 / WHAT WE DO HERE

Services tuned to ai, software & saas.

◈ 01
Patent prosecution
Software claims drafted to survive § 101 — improvement-to-computer-functioning, specific implementations, technical effects.
◈ 02
Trade-secret programs
Internal policies, employee training, NDA hygiene, and audit trails so you can prove misappropriation if needed.
◈ 03
Open-source audits
Dependency-tree license analysis, attribution generation, and remediation before acquisition.
◈ 04
Copyright registration
Source code, UI assets, marketing copy, and documentation registered with proper deposits.
◈ 05
AI/ML risk counseling
Training-data licensing, output-ownership terms, and customer-facing IP indemnities for foundation-model and RAG systems.
◈ 06
PTAB defense
Defending issued software patents in IPR proceedings — strategic claim-amendment and preliminary-response work.
// 04 / RELATED PRACTICES

Where this work lives.

Patents
VIEW PRACTICE →
Copyrights
VIEW PRACTICE →
PTAB
VIEW PRACTICE →
Litigation
VIEW PRACTICE →
// FAQ

Frequently asked questions about ai, software & saas IP.

  • What IP issues are unique to AI and software?
    AI and software companies face § 101 subject-matter eligibility risk on patents, open-source license obligations baked into modern dependency graphs, trade-secret considerations for model weights and training data, and trademarks that need to travel internationally because most software ships globally on day one. Lynch LLP coordinates § 101-resistant drafting, open-source audits, trade-secret programs, and international trademark strategy as a single engagement rather than discrete projects.
  • How does Lynch LLP approach filing timing for AI and software?
    AI and software companies face a faster disclosure cycle and § 101 subject-matter constraints that shape both timing and claim drafting. Lynch LLP aligns the disclosure pipeline with filing strategy and anchors claims to specific technical implementations, with continuations built into the roadmap.
  • How do we approach freedom-to-operate analysis in AI and software?
    AI/software FTO and open-source license review are interrelated and we run them together. The patent side focuses on patents covering specific system architectures — training pipelines, inference optimization, model orchestration, data flow — rather than the underlying models or general ML techniques. The OSS side identifies copyleft and reciprocal-license obligations that can affect your IP strategy: Apache 2.0 has explicit patent grants, GPL has implicit ones, and some licenses require disclosure of derivative works that conflict with patent strategy. We map the dependencies in your stack against both patent risk and license obligations, then deliver an assessment prioritized by impact: clear, monitor, mitigate, replace. For inference-heavy products we also analyze runtime IP exposure — patents on model partitioning, KV-cache optimization, and similar inference-time techniques are easy to overlook in a model-focused FTO.
  • How does our experience in AI and software differ from a general-practice firm?
    Three things distinguish our AI and software work. First, engineering depth — our attorneys came to law from electrical and mechanical engineering, and we've been drafting around §101 subject-matter eligibility since the Alice decision. We anchor claims to specific technical implementations rather than abstract ideas. Second, the attorney you meet is the attorney doing the work. Your patent strategy is set and executed by the same lawyer, not handed off to a rotating cast of staff. Third, continuation strategy is built into every roadmap. AI and software products iterate faster than most industries; the right pattern is to file the foundational application early, keep continuations pending, and add new claim sets as the product evolves.
◈ AI, SOFTWARE & SAAS IP

Bring us your ai, software & saas portfolio.

Book a consultation →View case studies