BriefCatch names AI chief to tackle fake legal citations

BriefCatch has appointed Josephine Petrick as Chief of AI Products as the legal-writing software company adds a citation-checking tool to its AI product line.

Petrick will lead development of AI tools designed to improve legal writing and help lawyers use the technology with greater confidence. Her arrival coincides with the rollout of RealityCheck, a feature that identifies hallucinated or unsupported legal citations before court filings are submitted.

The appointment highlights a wider issue in the legal sector, where growing interest in generative AI has been matched by concern over fabricated case references and other errors. Courts and law firms have faced increasing scrutiny over AI tools that can produce plausible but inaccurate legal material.

Petrick brings experience in both legal practice and product development. She clerked for the U.S. Court of Appeals for the Fifth Circuit and later became a partner at an appellate boutique, where her work included Supreme Court briefing and complex appeals.

She has also built legal technology products focused on rule tracking and case monitoring, including AI-based tools that use retrieval-augmented generation and related methods to improve the accuracy and reliability of legal research and analysis.

AI focus

BriefCatch is framing the hire around the need for legal AI products tailored to professional standards rather than general-purpose automation. The company develops software for legal professionals and integrates its editing tools into Microsoft Word.

Ross Guberman, founder and chief executive officer of BriefCatch, linked Petrick's legal background to the company's product plans.

"Josie combines deep appellate experience with hands-on expertise in building sophisticated legal technology," Guberman said.

He added: "As lawyers increasingly look to AI to support their work and boost their careers, her perspective will help us build products that are practical, trustworthy and grounded in the realities of legal practice."

BriefCatch has built its business around legal-writing assistance, with tools designed to improve clarity and precision in professional documents. RealityCheck extends that focus into verification by addressing one of the most closely watched risks in legal AI adoption.

That risk has become more prominent as lawyers experiment with generative systems for research, drafting and review. Hallucinated citations are a particular concern because they can enter filings in a form that appears authoritative unless checked against source material.

Legal standards

Petrick said the sector needs AI products shaped by a clear understanding of lawyers' professional obligations. Her comments point to a divide between general consumer AI systems and software designed for regulated professional work.

"Legal AI products need to be built by people who understand both the promise of the technology and the professional standards lawyers are required to meet," Petrick said.

She added: "BriefCatch has always stood out because it is grounded in real legal writing expertise, not generic automation. I look forward to helping build AI tools that support lawyers' judgment, improve their work and meet the level of trust this profession demands."

Petrick earned her J.D., magna cum laude, from the University of California, Hastings College of the Law, now UC Law San Francisco. Her appellate background is likely to be relevant for a company whose products sit close to the drafting and checking of legal submissions, where precision of language and authority is central.

Founded by Guberman, a legal-writing author and trainer of new federal judges, BriefCatch says its software is used by tens of thousands of law firms, courts, agencies and individual legal professionals. Its direct integration into lawyers' Word workflows has helped position it as a writing and editing specialist rather than a broad legal research platform.

The latest hire suggests BriefCatch sees the next phase of competition in legal software centering not only on drafting assistance but also on verification and trust. In legal practice, where unsupported authority can create professional and reputational risk, tools that confirm whether AI-generated citations are real may become as important as the systems that produce the first draft.