A human-authorship promise just met a partial AI admission

Every mainstream newsroom now carries an implicit authorship promise its contributors are quietly renegotiating in public. That is the real story behind UC Berkeley math professor Zvezdelina Stankova's admission that she used AI to help edit a 2,000-word San Francisco Standard op-ed arguing her students are five-to-eight years behind on fractions and basic algebra.

Berkeley sophomore Francis Luo, who covers tech for the Daily Californian, ran the piece through Pangram, an AI-detection tool, and got a 33% AI-generated score. Stankova replied that AI was used to 'help edit the piece' and that the work was 'the result of several hundred person-hours of intensive human work, of which about 80 hours are my own.' The Standard's response, that 'humans are behind every article we publish and take responsibility for every word,' treated the distinction as sufficient.

Most readers will read the admission as either a scandal or a non-story. Both miss the mechanism. The partial-disclosure gap—a characterization this story uses to frame the pattern—is the gap between a masthead-level human-authorship promise and a contributor's partial AI-editing admission, with no formal rule for where the line sits. Stankova's call to reinstate the SAT and the AI-in-academia fight are separate, legitimate debates the editing admission is now distracting from.

The next time an op-ed lands with AI fingerprints on it, the Standard's line is now being held up as the ruler. Outlets that want to keep the implicit promise will need to publish what 'help edit' means, where the line sits, and what disclosure looks like. The reader is now holding the yardstick whether the masthead wrote one down or not.