The water (or other equivalent reaction partner) is an important contributor to the energy calculus. In ATP, for instance, simply "breaking" a phosphoanhydride bond - say with imaginary molecular tweezers - by pulling off a phosphate would not be energetically favorable. We must therefore be careful not to say that breaking bonds in ATP is energetically favorable or that it "releases energy". Rather, we should be more specific, noting that the hydrolysis of the bond is energetically favorable.
An interesting example of how situational and context driven chemical and biological mechanisms can be. Isn't it unique and strange that the favorable mechanism here is strictly hydrolysis, which requires water, the most plentiful molecule in living things?