BAGHDAD (Reuters) - The U.S. military has released about 1,400 Iraqi detainees so far to mark the Muslim fasting month of Ramadan, after first making each take a pledge not to attack U.S. or Iraqi forces, a U.S. general said on Wednesday.
The vast majority of the 25,000 detainees held by U.S. forces are Sunni Arabs accused of involvement in the insurgency against the Shi’ite-led government and American troops.
Major-General Douglas Stone, the commander of U.S. detention facilities in Iraq, told reporters in Baghdad the programme of releasing about 50 detainees a day would likely continue after Ramadan ended this week. The holy month began in mid-September.
“We will continue to release every detainee who is no longer an imperative security risk,” he said.
Each of the prisoners released had been interviewed by a U.S. military panel to assess the risk they posed to Iraqi society. Those deemed to pose no risk were then required to make a pledge of good behaviour in front of an Iraqi judge.
“I want to thank the detainees ... who after going through this process have returned to Iraqi society and who themselves have chosen to build a better Iraq and have chosen to no longer engage in insurgent activity,” Stone said.
None of those who had gone through the programme had been re-arrested for attacking U.S. or Iraqi forces, he said. Sunni Arab leaders say many of those in U.S. detention are innocent of any crimes and are being held without charge.
Stone said there was a 65 percent conviction rate in cases referred to the Iraqi criminal courts.
The treatment of the detainees is an emotional issue for Sunni Arabs, and Iraq’s Sunni Arab vice president, Tareq al- Hashemi, has long called for their release.
The U.S. military reached a deal with Hashemi in August to conduct special Ramadan releases.
Stone said of the 25,000 in U.S. detention, 83 percent were Sunni and 16 percent Shi’ite. They included 860 aged 17 or under and 280 foreigners, mainly Syrians, Egyptians, Iranians and Saudi Arabians. Detainees were held about 300 days on average.