No, this is not an example of a biotech crop crossing with a weed. The “pig weed” in question here is actually what is known as Palmer Amaranth―a serious weed problem in its own right, regardless of herbicide tolerance. As with many other weeds that have become resistant to herbicides, including those long before biotechnology, it is not an issue of the tolerance moving to the weed by outcrossing, but rather that the use of the herbicide selects for resistant types that occur within the natural genetic diversity of the weed population. If anything, the issue has been that farmers have relied too heavily on Roundup and, in the case of this weed, have used it at lower rates that facilitated the adaptation of the resistant types over time. The best way to control weeds without selecting for resistance is to use different approaches over time to minimize that selection pressure. As for what herbicides are used to control the resistant types, they are not “highly toxic” in the sense that people imagine. They are toxic to the weed but generally not to animals, insects, etc. Weed control is, and always has been, a major challenge for farmers. Doing it with mechanical tillage mechanisms is problematic for the environment (erosion, loss of soil organic matter, high fuel consumption…). Herbicides, carefully and strategically used, are the best option.
QAbout regulation... Another answer by Steve Savage says the following -- "The USDA considers whether there are any “plant pest” issues with the specific crop and trait such as the ability to cross with weedy relatives." and "The EPA gets involved if