Our general editorial policy doesn't allow sources prior review of a full piece before publication, but we make a narrow, deliberate exception for researcher interviews: we do allow researchers to review the specific technical quotes we plan to attribute to them, solely to check for technical accuracy, before the piece runs.
Why we treat this category differently
Technical AI research often involves precise terminology where a small imprecision in how a researcher's spoken words get transcribed or summarized can meaningfully misstate a technical claim, in a way that's harder for a non-specialist editor to catch than an equivalent error in a more general business quote.
What this review is limited to
Researchers see only the specific quotes attributed to them, not the surrounding article, framing, or other sourcing, and their feedback is limited strictly to technical accuracy - not to softening a candid assessment or removing a quote they've simply reconsidered wanting attributed to them for other reasons.
Why this narrow exception doesn't compromise editorial independence
The distinction between reviewing for technical accuracy and reviewing for overall favorability is one we enforce carefully, and researchers who've asked for changes beyond correcting a technical imprecision have had those requests declined, with the original accurately transcribed quote retained instead.
Why we think this exception is worth making
The value of a technically precise quote from a genuine domain expert outweighs the small loss of strict pre-publication opacity this exception represents, particularly given how costly a technical misstatement can be to a piece's credibility with an informed audience.