A selection of my own posts. Below are short summaries; the links lead to the original posts. These publications are not automatically synchronised.
01 / AI & legal reliabilityThe Kribbenbijter judgment: why AI always needs checking
A familiar legal AI tool gave two different, incorrect interpretations of a classic Dutch contract-law judgment. Rob compared the answers with other tools and the original legal source. His conclusion: even a persuasively worded AI answer is not a verified legal answer.
Read the original LinkedIn post ↗02 / AI in litigationA polished AI-drafted claim is not necessarily a good claim
An AI-drafted pleading looked convincing, but its legal reasoning on default and directors' liability was inadequate. Rob explains why correcting such a document can take longer than starting again. Human review determines the quality of the result.
Read the original LinkedIn post ↗03 / Source research & current lawFinding a judgment AI does not know
Rob sought a recent disciplinary judgment. His legal AI tool offered general explanations but no verifiable source. Only independent searching uncovered the decision—a telling illustration of the importance of up-to-date research and independent legal judgment.
Read the original LinkedIn post ↗04 / Innovation & change managementAI in legal practice is largely about people
Following an event at the Dutch Supreme Court on AI adoption in law firms, Rob reflects on the need to understand processes, engage people and test systems carefully. Introducing technology is not the same as using it wisely.
Read the original LinkedIn post ↗05 / AI & family lawAI in family law: opportunities and sensitivities
A conversation with a family-law colleague raised questions about adopting AI in a field where privacy, emotions, children and vulnerability are central. AI calls for care suited to each case.
Read the original LinkedIn post ↗