Afghan leaders are pressuring their American counterparts to return much of the air and intelligence (surveillance and analysis) support that was largely withdrawn by the end of 2014. Actually, the decline was somewhat gradual but after 2014 the Taliban and drug gangs sensed that this support had reached such a low level that a general offensive could be undertaken successfully in 2015. The Afghan forces, even without American air and intel support, fought back more effectively than the enemy expected but the offensive continues into 2016. While the police and army losses are higher the civilian deaths are much higher and that has persuaded the United States to approve the return of some of the air and intel support. Some, but not enough.
To put it into perspective consider that in 2011 U.S. warplanes carried out 346,000 combat sorties in but only 35,000 of these sorties resulted in the enemy being bombed or shot at by the aircraft. In 2012 the number of sorties where the aircraft attacked dropped to 29,000, 22,000 in 2013, 13,000 in 2014 and 4,700 in 2015. At the current rate there will be less than 500 of these sorties in 2016. The U.S. policy of declaring victory and going home is not working. So there are plans to increase the number of combat sorties in 2016 but so far no action. Part of the reason for no action is the need for more American Special Forces on the ground to act as air controllers (to verify the target and actually call in the air attack) and the need for more aircraft to carry out surveillance and reconnaissance to provide better intel on what the enemy is up to. Then there is the increased pressure to get Afghan forces to observe American level ROE (Rules of Engagement). The problem here is that the Afghan are more tolerant of civilian losses in a battle but that is unacceptable to American leaders and when American air strikes end up causing the civilian casualties the Americans, not the Afghans calling for the air support, get blamed.