In the context of energy saving and emission reduction to protect the environment has become a social consensus, environmental benefit is a social responsibility that enterprises must actively undertake, and the technological advantages of information technology companies make them more equipped to enhance environmental benefits. Therefore, this paper proposes countermeasures for information technology companies to enhance their social responsibility and environmental benefits. To achieve practical countermeasures for energy savings and emission reduction, this paper proposes a method to improve the LSTM power prediction model using the Sparrow search algorithm for the company’s electricity management. The model optimizes the network structure of traditional LSTM by searching for parameters such as learning rate, iteration number, and the number of neurons in the two hidden layers of LSTM through the sparrow fitness function. The model proposed in this paper has an average prediction error rate of 3.32% in predicting the annual electricity consumption of a county for five years from 2018 to 2022. After an enterprise introduced the social responsibility practice countermeasures proposed in this paper, the company’s net profit margin increased by 12.2% on average, CO2 emission reduction was 963.2 tons, the average monthly electricity consumption was reduced by 24936.5 kWh, and the environmental performance score increased from 3.2 to 5.4.