Our approach to covering individual AI companies has evolved since we started, in ways we think are worth stating explicitly.
We stopped treating executive interviews as inherently favorable coverage
Early on, securing an executive interview sometimes meant the resulting piece leaned more favorable by default. We've worked to separate access from favorability, applying the same scrutiny regardless of whether a company cooperated.
We got more disciplined about distinguishing company narrative from verified fact
We now more consistently flag when a claim comes from company messaging versus independent verification, a distinction earlier coverage sometimes blurred.
We started tracking company claims over time
We didn't systematically revisit whether past company claims about roadmaps or capabilities actually materialized. We now do this more deliberately, since a company's track record on past claims is genuinely relevant context for evaluating new ones.
Why we're naming this evolution
Company coverage is particularly susceptible to access-driven bias, and we think being explicit about how we've tightened our own practices matters more than simply asserting independence.