The lakes that form via ice‐rich permafrost thaw emit CH4 and CO2 to the atmosphere from previously frozen ancient permafrost sources. Despite this potential to positively feedback to climate change, lake carbon emission sources are not well understood on whole‐lake scales, complicating upscaling. In this study, we used observations of radiocarbon (14C) and stable carbon (13C) isotopes in the summer and winter dissolved CH4 and CO2 pools, ebullition‐CH4, and multiple independent mass balance approaches to characterize whole‐lake emission sources and apportion annual emission pathways. Observations focused on five lakes with variable thermokarst in interior Alaska. The 14C age of discrete ebullition‐CH4 seeps ranged from 395 ± 15 to 28,240 ± 150 YBP across all study lakes; however, dissolved 14CH4 was younger than 4,730 YBP. In the primary study lake, Goldstream L., the integrated whole‐lake 14C age of ebullition‐CH4, as determined by three different approaches, ranged from 3,290 to 6,740 YBP. A new dissolved‐14C‐CH4‐based approach to estimating ebullition 14C age and flux showed close agreement to previous ice‐bubble surveys and bubble‐trap flux estimates. Differences in open water versus ice‐covered dissolved gas concentrations and their 14C and 13C isotopes revealed the influence of winter ice trapping and forcing ebullition‐CH4 into the underlying water column, where it comprised 50% of the total dissolved CH4 pool by the end of winter. Across the study lakes, we found a relationship between the whole‐lake 14C age of dissolved CH4 and CO2 and the extent of active thermokarst, representing a positive feedback system that is sensitive to climate warming.