This paper introduces learner model negotiation not only as a means to increase the accuracy of the learner model and promote metacognitive activities as in past examples, but also as a way to help learners correct peer assessment entries in their learner model, that they consider inaccurate. While open learner models are not new, and negotiated learner models have been developed before, in today’s learning contexts of potentially big data from many sources including other learners, some kind of approach to managing the data as well as helping learners to understand and accept it, or correct it, is needed.