As questions around AI safety, evaluation and global standards grow, our Chief Product Officer Nur Hamdan puts the problem simply: “Nobody should grade their own exam.”
If we want AI we can trust, how we measure it matters just as much as how we build it.