Product, Commercial & IP Counsel · Generative AI and Creative Tech

Counsel for the companies building what’s next.

Liam J. Penwell is a California-licensed attorney who serves as outside product, commercial, and IP counsel to 30+ technology and creative companies across Silicon Valley and the Bay Area. His work runs the length of the product lifecycle: platform and developer terms, SaaS and content licensing, privacy under CCPA/CPRA and GDPR, and the generative-AI questions that come with shipping a product: training-data rights, ownership of model outputs, and synthetic voice and likeness. Founders also bring him the company itself: entity formation and governance, equity and investor agreements, and the securities questions that come with a first raise. When a matter turns adversarial, he carries it himself, having served as lead counsel in federal administrative hearings. He carries deal-side regulatory experience from inside SoftBank Investment Advisers, where he owned the compliance sign-off process for the investment deal stream under the Investment Advisers Act and FCA rules. Direct, senior-level counsel that clears the path from build to launch.

Get in Touch

Liam J. Penwell · California-licensed attorney · San Francisco Bay Area

Six places the practice goes deep.

Product & Commercial Counsel

The contracts that move a startup's revenue are also the ones that decide its exposure two rounds later, and they rarely get the senior attention they deserve. Penwell drafts and negotiates SaaS and platform terms, API and data-licensing provisions, work-for-hire and content-rights agreements, and the MSAs, order forms, and security and data terms that unlock enterprise sales. First-line counsel to product and engineering teams, translating legal requirements into guidance a team can build against rather than route around. For consumer AI products: output-reliance disclaimers, acceptable-use terms, age-gating, and non-affiliation language, written for companies shipping models to the public.

ii.

Privacy & Data Protection

The hard privacy question for an AI company is not the policy text, it is what the model was trained on and what the company can defend in front of a regulator. Penwell handles privacy as a product question: data mapping, data processing agreements, consent and notice frameworks, DPIAs, and cross-border transfer mechanics across CCPA/CPRA, GDPR, and the widening patchwork of U.S. state privacy statutes. For machine-learning products that means the data side of training and inference, not just the public-facing policy: provenance of training sets, secondary-use limits, and automated-decision and profiling obligations now landing under CPRA rulemaking and the state laws following it.

iii.

Regulatory & AI Governance

A model going to market now meets a regulatory map that did not exist three years ago, and enterprise buyers and investors expect the company to have read it. Penwell builds the governance layer they ask for: AI use policies, model and vendor risk assessments, and audit-ready documentation mapped to the NIST AI Risk Management Framework. The work covers EU AI Act risk classification and general-purpose AI obligations, the multiplying U.S. state AI and transparency statutes, and FTC consumer-protection exposure around AI claims. That regulatory instinct was formed inside SoftBank Investment Advisers, clearing deals to close under the Investment Advisers Act, Securities Act, Exchange Act, Dodd-Frank, and FCA rules.

iv.

IP & Content Rights

The copyright questions that used to live in entertainment law now sit at the center of building an AI product. Penwell works where they collide: fair-use analysis for model training, training-data and catalog licensing, ownership of model outputs, and the synthetic-voice and digital-replica questions emerging under state right-of-publicity statutes. The practice runs ownership and enforcement end to end: content provenance, infringement exposure, DMCA Section 512 notice-and-takedown strategy, and the cease-and-desist and settlement work that resolves disputes before they reach a courtroom. Grounding the analysis is two decades on the creator side of licensing, where these rights were negotiated long before they became model-training questions.

Trust & Safety

Trust and safety frameworks work best designed in from the start, scaled for the day real users arrive. Penwell drafts the acceptable-use policies, community guidelines, and content-moderation and escalation frameworks platforms operate at scale: UGC governance, DMCA notice-and-takedown, transparency reporting, and the user-notice and risk-assessment duties landing under the EU Digital Services Act. For generative-AI products this is where rights, output safety, and product design meet, and where the escalation path for sensitive content has to be designed deliberately rather than improvised under pressure.

vi.

Music & Entertainment

Two decades inside the music industry sit behind this practice, from studio floor to licensing table to counsel. Penwell advises artists, labels, and creative businesses on recording and publishing agreements, performance contracts, synchronization and catalog licensing, and royalty and rights management. The same grounding informs the AI-era version of these questions: licensing of sound recordings for training, synthetic-voice right-of-publicity exposure, and ownership of AI-assisted output. That experience was earned, not studied: founder of Diamond Sonic, a San Francisco production studio whose clients included Google, Uber, LinkedIn, Salesforce, and Twitter.

The whole story

Experience & Credentials

At SoftBank Investment Advisers, he owned regulatory diligence and compliance sign-off for the investment and divestment deal stream. He advised deal teams and senior leadership under the Investment Advisers Act, FCA rules, and LatAm and APAC obligations, reviewing and negotiating subscription agreements, side letters, and co-investor NDAs, and helping build the firm’s capital-markets compliance process, information-barrier protocols, and regulatory filing checklists from the ground up. Over time his responsibility expanded to span the firm’s Latin America transactions. The work taught him how legal review actually functions inside a large, fast-moving, heavily regulated organization.

Today he runs Penwell Law as embedded outside counsel to founders across tech and the creative industries, bridging legal and business judgment on every decision. He works best where the technology moves quickly and the legal framework is still taking shape. He has served as lead counsel in federal administrative hearings before Administrative Law Judges. He has also taught and lectured for fifteen years. Most recently that meant a Music Law and Business course at the San Francisco Conservatory of Music, built around copyright, trademark, contracts, and AI-and-copyright issues. Earlier it meant music production and music business at San Francisco State University, along with workshops and online courses. As Liam Shy, he released original music featured in Microsoft and Insomniac Games' Sunset Overdrive, for Xbox. As Mountain Of Youth, he performed at Coachella in 2016. And before any of it, at nineteen, he wrote and passed a municipal law, signed by the Mayor of San Francisco, then Governor of California, Gavin Newsom.

Liam came to law from the music industry. For close to two decades he worked as a producer, studio owner, and event operator. He founded Diamond Sonic, a San Francisco audiovisual production company and recording studio whose clients included Google, Uber, LinkedIn, Salesforce, and Twitter. Creative-AI products run on a particular set of questions: who owns a model's output, what training data a company can use, when a voice or a likeness is protected. He understood those questions from the creator's side long before he argued them from the lawyer's. He has sat in every seat at that table: the artist whose work is used, the technologist building the product, and the lawyer allocating the rights and risk.

J.D., Northwestern California University School of Law — 2021
Admitted, State Bar of California — 2021 · Bar No. 339157
GenAI for the Legal Profession: Power User Edition, Berkeley Law Executive Education — 2026
Pyramind Institute, Music Business and Production — 2008

Common questions

Questions, answered.

Who is Liam J. Penwell?

Liam J. Penwell is a California-licensed attorney (State Bar No. 339157) serving as outside product, commercial, and IP counsel to 30+ technology and creative companies across Silicon Valley and the Bay Area. His background spans deal-side compliance inside SoftBank Investment Advisers and two decades in the music industry.

What does outside product counsel do for an AI startup?

Product counsel handles what the company ships: platform and developer terms, SaaS and data licensing, privacy under CCPA/CPRA and GDPR, and the training-data, model-output, and voice-and-likeness questions specific to generative AI. Guidance a team can build against, clearing the path from build to launch.

Does Penwell Law work with early-stage startups?

Yes. Founders bring the company itself: entity formation and governance, equity and investor agreements, and the securities questions that come with a first raise. Engagements run from formation through scale.

What industries does Penwell Law serve?

Technology and creative companies: generative-AI products, platforms and marketplaces, consumer applications, and music and entertainment businesses, from solo founders to funded startups.

Can Penwell Law advise on the EU AI Act and GDPR?

Yes, as they bear on U.S. companies: EU AI Act risk classification and general-purpose AI obligations, GDPR data protection, and DSA-era trust-and-safety duties, counseled as compliance exposure for products shipping from California.

Where is Penwell Law located?

The practice is based in Burlingame, California, in the San Francisco Bay Area, and works with clients across Silicon Valley and nationally. California and federal practice.