By Nick Robertson (January 2, 2026, 21:24 GMT | Comment) — As companies increasingly customize AI models using proprietary software, agents, and data labeling frameworks, Saul Ewing partner Matthew Kochel tells MLex that IP practitioners must remain aware of commercially valuable trade secrets in data filtering and dataset creation. While the trade secret implications of AI training datasets and models have received significant attention from IP practitioners, custom tools used to filter and manage data within proprietary models are an overlooked trade secret protection opportunity, Saul Ewing partner Matthew Kochel told MLex.
Prepare now for tomorrow’s regulatory changes
MLex identifies where risks to your business occur, with expert reporters from around the world providing exclusive news and in-depth analysis of proposals, investigations, enforcement actions, and decisions that matter to organizations and clients now and in the long term.
Know what others in the room don't know with features like:
- Daily newsletters on antitrust, M&A, trade, data privacy and security, technology, AI and more
- Custom alerts for specific filters such as region, industry, topic, or company to suit your practice needs
- Predictive analysis from expert journalists in North America, the UK, Europe, Latin America and Asia Pacific
- Hand-picked case files with news, analysis, and source documents in a single timeline
Experience MLex today with a 14-day free trial.
