Protecting your IP in the age of generative AI

Phoenix companies are racing to adopt generative AI, but many haven't considered how everyday AI use can quietly erode the intellectual property that defines their competitive edge.

Why it's important: For most Arizona companies, intellectual property — from proprietary formulas to source code to product roadmaps — is often what differentiates a company in the market.

• When employees input sensitive details into AI systems, especially consumer-grade tools, those actions may later be scrutinized in trade secret disputes or patent filings.

The challenge: Most company-wide AI use hasn't caught up with a formal governance framework, and boards that routinely oversee cybersecurity and data privacy now need to treat AI governance with the same strategic seriousness.

Fennemore, one of the nation's fastest-growing business law firms, is seeing this risk play out in manufacturing, health care, real estate, and technology companies across Arizona, especially as companies scale.

Why now: Generative AI is now embedded into workflow processes across industries, companies, and departments.

• The speed and efficiency gains are real, but so is the legal exposure that comes with unstructured use.

What you need to know: Many consumer-grade AI tools may retain or use input data to improve their models unless users actively opt out.

• Meanwhile, individual-tier plans don't always include confidentiality agreements that enterprise plans do.

From the perspective of U.S. patent law, that distinction matters.

Here's the deal: Courts have long assessed how companies handle sensitive information when evaluating trade secret claims. If proprietary material is shared without adequate protection, a company's ability to claim trade secret status can weaken.

Patent law presents another risk. In general, public disclosure of an invention more than one year before filing can bar patent protection.

• Whether entering information into an AI system constitutes "public disclosure" is still evolving, and will likely depend on contractual protections, the platform used and the specific facts of each case.

The strategy: Mario Vasta, an intellectual property attorney at Fennemore, frames it plainly: if you wouldn't post it publicly, don't paste it into a consumer-grade chatbot.

Leadership teams should also consider the following approaches to assess and address their disclosure risk:

• Review AI platform agreements carefully. Understand what consumer and enterprise plans actually provide, and opt out of data-sharing and model training where possible.
• Build an internal AI policy that defines who can use which tools for what tasks, and what categories of information are off-limits.
• Train your team with explicit guidance on what should never be entered into AI systems, rather than a general credence that "sensitive stuff" should stay internal.
• If you plan to patent it, file early. A provisional application is often far less expensive than litigating over what counted as disclosure.

Take note: Fennemore's intellectual property team helps companies identify gaps, assess risk and design practical safeguards that support innovation while preserving long-term protection.

The takeaway: AI can increase speed, creativity and operational efficiency, but protecting a competitive advantage depends on protecting the information that makes a company distinct.

• Leadership teams that work with experienced IP counsel to treat AI use as a governance and legal matter will be better positioned to defend their assets if they're ever challenged.

Evaluate your AI governance framework.