Could you detail more ? Do you compare one C=O with two C-O or with two C=O? Are the carbons tetravalent?
Caution also: bond energies vary a lot, often too much to be usable. They tend to be mean values. Long ago, some people tried to make very complicated rules to improve the accuracy, including the effects of all neighbour bonds and cycles, but it never sufficed for me. As an example, two C-O on the same C differ from neighbour C that differ from distant C.
Just in case you compared the heats of formation of CO and CO2, the second double bond is weaker than the first one, but heats of formation don't reflect it immediately because the heat of formation of graphite is taken as zero despite it contains many bonds. That is, the formation of CO must break all C-C bonds, which reduces the produced heat, while from CO to CO2 the carbons are already separated. The O-O bonds act identically on both reaction heats.
You could check N-N, N=N and N≡N (was that the question?): the energy per shared pair increases much from single to triple bond.