LiCoO2 (LCO) is a widely used cathode material for lithium-ion batteries with superior volumetric energy density and tap density. Traditionally, only about 50% of the total lithium content are utilized in practical applications since it is believed that more than 50% Li extraction can cause irreversible reactions and structural degradation with oxygen releases from the lattice. However, there has always been a great interest in going beyond the delithiation level of x=0.5 in Li1-xCoO2 to utilize the full potential of its energy density. Therefore, a clarification of the controversial issues related with the oxygen activities in highly delithiated LCO becomes critical for guiding the development of high energy density LCOs . In this work, we combined the results of several modern experimental techniques, especially resonant inelastic X-ray scattering (RIXS) and neutron scattering pair distribution function analysis (nPDF), together with the theoretical calculations to study the nature of the oxygen activities at highly delithiated state of LCO. Our results conclude that the lattice-oxygen-redox nature of LCO, which means the oxygen redox takes place globally in lattice, are quite different from the oxygen oxidation through dimerization reported in literature. Our RIXS results directly reveal the reversible lattice oxygen redox, and nPDF results show that the O-O pair distance is shortened from 2.63 Å for pristine state to 2.47 Å for highly delithiated state of LCO. Pairs with similar short length were previously associated with O-O bond formation in 4d/5d systems. Strikingly, although with the shortened O-O distance, our theoretical calculations show that no O-O bonding is formed in LCO, in sharp contrast with the calculation results of 4d/5d transition-metal oxides (Ru and Ir), as well as 3d Li-rich layered systems, based on exactly the same theoretical calculation method. These new results challenge the widely accepted O-O bond formation being the natural result of oxygen redox in highly delithiated LCO system. Our combined structural, spectroscopic and theoretical results clearly reveal that, while the lattice-oxygen-oxidation takes place in LCO, the broadly accepted O-O bonding is not a necessary condition. In addition, our findings indicate that oxygen release is not necessarily the intrinsic channel of oxidized oxygen in LCO at highly charged states, providing the rationality for achieving a reversible deep delithiation and high energy density in LCO based electrodes. Acknowledgements The work done at Brookhaven National Laboratory was supported by the Assistant Secretary for Energy Efficiency and Renewable Energy, Vehicle Technology Office of the U.S. Department of Energy through the Advanced Battery Materials Research (BMR) Program under contract DE-SC0012704.