Infactory’s use of generative AI while the technology is white hot no doubt piqued investor interest. So, too, did the promise of addressing the very real problem of online fact-checking.

The young company is quick to point out the ways in which large language models (LLMs) won’t be utilized on its platform. Specifically, they will be used to determine the intent of natural language search queries, rather than the results themselves. After all, the black box nature and built-in biases of large models run counter to the point of a fact-checking service. Questionable Gemini search results are clear evidence of this phenomenon.