It might help to distinguish genetic engineering from traditional techniques for producing food.
Humans have been selectively breeding plants and animals for tens of thousands of years to get certain desired traits. Over time, for example, farmers (and scientists) have bred corn to become larger, to hold more kernels on an ear, and to flourish in different climates. That process has certainly altered corn’s genes. But it’s not usually considered “genetic engineering.”
Genetic engineering, by contrast, involves the direct manipulation of DNA, and only really became possible in the 1970s. It often takes two different forms: There’s ”cisgenesis,” which involves directly swapping genes between two organisms that could otherwise breed — say, from wheat to wheat. Or there’s ”transgenesis,” which involves taking well-characterized genes from a different species (say, bacteria) and transplanting them into a crop (such as corn) to produce certain desired traits.
Ultimately, genetic engineering tries to accomplish the same goals as traditional breeding — create plants and animals with desired characteristics. But genetic engineering allows even more fine-tuning. It can be faster than traditional breeding, and it allows engineers to transfer specific genes from one species to another. In theory, that allows for a much greater array of traits.
Here’s a diagram from the Food and Drug Administration that illustrates the two methods: