Patents can be obtained by inventors who have not implemented the related idea in practice or physically demonstrated that it works; rather, the patents are obtained on the basis of 'fictional' evaluations and data - a fact likely not well appreciated by most scientists who use them as technical records. In a Policy Forum, Janet Freilich and Lisa Larrimore Ouellette highlight the common practice of including "prophetic" examples in patents - particularly in the fields of chemistry and biology, where patents routinely describe the outcomes of experiments that have not been conducted - and suggest labels in patents, to better call out such examples. According to patent law, as long as hypothetical examples are not written in the past tense, they can be included in a patent's language and used to expand its legal coverage beyond the specifics of what has been done in the lab, including to block competing technologies. Despite the long history of the practice, Freilich and Larrimore Ouellette argue that the common use of prophetic examples in patents has the great potential to sow confusion; this is particularly likely among those who may not be aware of the practice. Because the legal patent language required of prophetic examples mimics real experiments and often includes excessive detail concerning samples or numerical results, language describing these examples is likely to mislead non-legally minded readers, like scientists. In a survey of 100 randomly selected patents that were cited in scientific literature and featured only prophetic examples, 99 were not cited in a way that made clear they contained no actual data, show the authors. Freilich and Larrimore Ouellette suggest that prophetic examples in new patent applications be labeled to make a clearer distinction between prophetic and non-prophetic examples. Furthermore, patent drafters should be mindful to avoid potentially confusing details or language.