The proposed high-dorsiflexion assistive robotic technology (RT) is aimed to assist ankle movement during gait rehabilitation. A McKibben-type artificial muscle facilitates dorsiflexion in the swing phase by lifting a forefoot of a stroke patient's paralyzed side. In spite of the great potential of improving dorsiflexion owing to the proposed RT observed in the previous clinical results, the current approach is applicable only to the passive ankle gait rehabilitation with full assistance. It is believed that ensuring patients' voluntary effort in rehabilitation is the key for long-term recovery. In the present research, we conducted an experiment to evaluate our RT's effects on six healthy participants, and analyzed whether their voluntary ankle movements were affected due to the current assistance strategy. Compared with normal walk, significant decrease in processed surface electromyography signal (p = 8.62×10-3), significant early timing of a push-off motion (p =2.60×10-4), and significant smaller push-off angle (p = 1.50×10-4) were observed compared with intervention of the proposed RT. On the other hand, a significant trend of decrease in the ankle dorsiflexion rate in the swing phase is observed during intervention (p = 0.06949). Based on the obtained results, we concluded that unlike the contraction speed of an artificial muscle that did not show any considerable impact, the excessively early activation of the RT assistance affected voluntary ankle movement.