Abstract: Unmeasured confounding remains one of the most significant threats to the credibility of findings from observational studies. Recent developments in negative control methods, also known as proximal causal inference, offer promising strategies to strengthen causal conclusions. These approaches leverage negative controls —variables that have no direct causal relationship with either the exposure or the outcome — to detect and adjust for unmeasured confounding. In this talk, I will review the foundations of negative control methods and introduce the double negative control framework. I will then present our recent work extending this framework to settings where some candidate negative control variables may themselves be invalid. I will conclude with a discussion of open challenges and future research directions.