The U.S. and the Taliban signed the “Agreement for Bringing Peace to Afghanistan” (Agreement) on February 29, which sets into motion a series of steps to wind down the U.S. war in Afghanistan. As some have already suggested, the Agreement asks for, and accordingly seems likely to get, rather little. The Agreement makes clear that U.S. forces are to withdraw but makes remarkably few demands on the Taliban to support a peaceful resolution in exchange for the U.S.’s commitments. Less appreciated is that these same features also mean that it is difficult to conclude — even as significant violence resumes in Afghanistan and some have called for U.S. action cutting against the Agreement — that either side is already violating the Agreement’s stated terms.

This post dissects the Agreement and its legal implications, addressing four questions: 1) what does the Agreement commit each side to do? 2) is the Agreement a ceasefire, peace agreement, or neither? 3) is the Agreement legally binding? and 4) what might the Agreement mean for the end of the armed conflict and ongoing detentions?

1) What does the Agreement Actually Require of Each Side?

According to the text of the Agreement, the most concrete and onerous commitments fall on the U.S., which has agreed to a timeline for withdrawal, return of over 5,000 political Taliban prisoners, and review of sanctions imposed on Taliban members. In exchange, the Taliban has promised to participate in intra-Afghan dialogues that began on March 10 and to not threaten the security of the U.S. and its allies. However, the Agreement’s details (and lack thereof) are key to understanding the U.S.’s withdrawal commitments, and both side’s subsequent uses of force.

The Agreement opens by explaining that a “comprehensive peace agreement is made of four parts.” As listed, these four parts are:

  1. guarantees and enforcement mechanisms to prevent Afghanistan territory from being used against the security of the US and its allies;

  2. guarantees, enforcement mechanisms, and a timeline for U.S. withdrawal;

  3. intra-Afghan dialogues beginning March 10, 2020; and

  4. that “permanent and comprehensive ceasefire will be an item on the agenda” of this dialogue.

The Agreement suggests that these parts are “interrelated” and will be “implemented in accordance with its own agreed terms.” It then suggests that “[a]greement on the first two parts paves the way for the last two parts.”

“Interrelation” between the parts of this plan would seem to make sense. Years of thinking has gone into the issue of how best to ensure U.S. interests upon withdrawal, and any peace agreement will likely come in parts. But what is striking is that while the Agreement suggests that the first two parts (agreement on the Taliban’s security efforts and the timeline for U.S. withdrawal) serve as conditions for the last two parts (dialogues and discussion of a permanent ceasefire), such conditions are not said to run in the reverse. This means that the U.S. withdrawal may happen (or potentially may even be expected to happen) whether or not a permanent ceasefire or other outcome is reached at the intra-Afghan dialogue, the critical forum for future Taliban-Afghan relations, including any potential protection for women’s rights.

As to the specifics of the drawdown, the Agreement makes clear that the U.S. is leaving, again with seemingly little contingency. The Agreement provides that the U.S. is committed to reducing its forces to 8,500 within 135 days, a plan apparently long in the works. The U.S. is further “committed” to completely withdraw within 14 months.

Some read more contingency into the withdrawal by focusing on the provision pertaining to drawdown after the first 135 days, which suggests such will happen “with the [Taliban’s] commitment and action on [its] obligations.” These obligations are as follows:

  1. “not allow any of its members, other individuals or groups, including al-Qa’ida, to use the soil of Afghanistan to threaten the security of the United States and its allies”;

  2. send a message to its members “not to cooperate with groups or individuals threatening the security of the United States and its allies”;

  3. “prevent any group or individual in Afghanistan from threatening the security of the United States and its allies,” and “prevent them from recruiting, training, and fundraising” and “not host them”;

  4. “deal with those seeking asylum or residence” in accordance with international migration law; and

  5. not allow into Afghanistan or “provide visas, passports, travel permits, or other legal documents to those who pose a threat to the security of the United States and its allies.”

Whether the overall 14-month timeline, or just the drawdown, is tethered to the Taliban’s compliance, is not explained in the text, and differing views appear to have emerged. A corresponding joint declaration signed between the U.S. and Afghanistan on February 29 notably seems to provide more flexibility than the Agreement with the Taliban. It suggests that the 14 month withdrawal will take place “[c]onsistent with the joint assessment and determination between the United States and the Islamic Republic of Afghanistan.”

2) A Ceasefire, Peace agreement, or Neither?

Likely neither. While extensive academic and practical debate exists over the use of such terms, the Agreement lacks critical features of either a ceasefire, or a peace agreement, as they are more commonly understood and defined.

The Agreement emerged out of a two-week “partial truce,” which saw a reduction in violence as a confidence-building measure for its signing. But as seen in the text, the “partial truce” did not make its way into obligations imposed on either side. Reports suggest that a ceasefire had become a nonstarter for the Taliban long before February. Others explain that while a general violence-reduction provision had been considered, it also didn’t make the cut, leading to a gap in expectations about the meaning of the Agreement.

Confusion seems to have nonetheless surrounded the Taliban’s obligations. The Agreement provides that the Taliban commits to not allow its members or territory to be used to threaten the “United States and its allies,” though it does not state whether Afghanistan is one of these “allies.” The reference to “Afghanistan” in other parts of these obligations, combined with reference to classified “enforcement mechanisms” to ensure these obligations, casts doubt on the view that the Taliban has a general obligation to reduce violence, and instead seems to suggest that the Taliban is committing instead solely to preventing global terrorism. Classified enforcement mechanisms would hardly seem necessary to verify whether the Taliban has prevented routine attacks on Afghan forces and civilians.

On the other hand, some scholars envision ceasefires as a “a first link in a chain running from war to peace,” and as “a preliminary and provisional step, providing a breathing space for the negotiation of more lasting agreements.” This is what the Agreement seems to be. But uncharacteristic of many ceasefire agreements, it does not provide any timing or other parameters for halting uses of force. The Agreement also mentions “peace” and the obligation to put a “comprehensive peace agreement,” on the agenda of dialogues. Of course it does not yet provide a permanent ceasefire, or a plan for future Taliban-Afghan relations, though it is framed as part of an overall process. On the other hand, it is notably not clear how even the U.S. plans to square with the Agreement its own obligations under the U.S.-Afghanistan bilateral security agreement with respect to the use of force, particularly as the joint declaration “re-affirms” the United States’ security obligations and “readiness to continue to conduct military operations in Afghanistan” for purposes of counter-terrorism. It is also noteworthy that the joint declaration states that the third part of the peace plan is “political settlement” emerging from the intra-Afghan dialogue, not that such is merely an “agenda” item, as in the language of the Taliban agreement.

3) What is the Legal Force of the Agreement?

The Agreement likely can’t be considered a “treaty” in the traditional sense, but it may well be binding. The Agreement refers to the Taliban as the “Islamic Emirate of Afghanistan which is not recognized by the United States as a state and is known as the Taliban.” Article 2 of the Vienna Convention on the Law of Treaties (VCLT) provides that a treaty is “an international agreement concluded between States in written form and governed by international law.” However, Article 3 of the VCLT provides that the Convention does not affect agreements concluded between States and “other subjects of international law.”

Commentators have expressed differing positions on whether agreements between States and armed groups are binding under customary international law, given that such groups are typically understood not to possess full international legal personality. Some haltingly answer this question in the negative, but suggest the inquiry is highly fact and agreement specific. More agree that such agreements are binding but differ as to why.

Supporting the binding view, it is undisputed that individuals and armed groups incur obligations under international humanitarian law (IHL), and (increasingly) under human rights law. Common Article 3 of the Geneva Conventions also envisions the possibility of binding agreements with non-State actors and encourages parties to a conflict to conclude agreements concerning IHL with non-State actors. While Common Article 3 does not specify whether agreements with armed groups create obligations rooted in international law, domestic law, or some sort of sui generis regime, some argue that when a State is a party to such an agreement, this fact elevates such agreement to the international order, even though Article 3 makes clear that such agreements cannot influence legal personality. Another view suggests looking to the content of the obligations and whether they are international in nature.

The few courts to consider agreements between States and armed actors have tended to find them not binding under international law, but likely binding under domestic law. In a 2004 decision that has attracted much criticism, the Special Court of Sierra Leone Appeals Chamber held in Prosecutor v. Kallon and Kamara that the Lomé Agreement between the Government of Sierra Leone and the Revolutionary United Front was not a treaty because it was concluded with a non-State group. The Constitutional Court of Colombia reached a similar conclusion in 1995. It is unclear what position the U.S. would take on the issue, and the U.S. Department of Defense Law of War Manual offers little guidance, apart from potentially suggesting the relevance of “good faith.”

Third-party guarantors or enforcers play an important role in ensuring agreements between States and armed groups on the international level, and in evidencing the potential international nature of obligations contained in such agreements in the first place. Significant efforts were made in the peace agreement between Colombia and the FARC, for example, to elevate that agreement to the international level and to ensure its enforcement. As Laura Restrepo explains, it was deposited in Switzerland, the Colombian government made a unilateral declaration before the United Nations Secretary General, and the U.N. Security Council was asked to adopt it.

On March 10, the U.N. Security Council welcomed the Taliban agreement and the joint declaration as significant steps towards ending the war and opening negotiations. The Council called on both sides to advance the peace process and to pursue additional confidence-building measures in “good faith,” while emphasizing the importance of including women and minorities in any political settlement. However, it is difficult to find much in the resolution to conclude that the Taliban agreement is itself binding, or that the resolution requires much by way of concrete compliance. The Security Council has previously expressed support for the signing of ceasefire and peace agreements, endorsed agreements arrived at, and in some cases adopted coercive measures to ensure their compliance (e.g. Liberia (1992) and Sierra Leone (1998)). At other times, according to Olivier Corten and Pierre Klein, the Council has even adopted language suggesting that such agreements are themselves binding. Still, according to Corten and Klein, the Council has not typically indicated that traditional forms of international legal responsibility flows from a breach.

In any event, the role of the United States in the Security Council should be remembered in relation to the possibility of more forceful measures (particularly with respect to the obligations of the U.S. itself), as well as the potential impact any position taken by the United States on agreements with non-State actors may have on future negotiations.

4) What About Guantánamo Detainees?

Finally, it is worth considering what the Agreement might mean for obligations pertaining to the end of armed conflict. Just Security readers will recall several posts (see here, here, and here) prompted by the United States’ end of Afghanistan combat operations in 2015. A number of theories exist as to when non-international armed conflicts (NIACs) conclude, including when the criteria used to establish the initial existence of a NIAC are no longer present, when combat operations expire, when a resumption of hostilities is no longer probable, or (less commonly) when a peace settlement is reached (see the International Criminal Court Trial Chamber Judgment in Prosecutor v. Bemba). Afghanistan presents a tricky situation requiring the parsing of what arguably are (or have been) several different armed conflicts involving different players (the Taliban, Al-Qaeda, Afghan government, etc.). If the U.S. adheres to the terms of the Agreement and withdraws its forces, that may well mark the end of its participation in armed conflict against the Taliban and potentially (though less clearly) Al-Qaeda in Afghanistan.

In need of revisiting as the Afghanistan process unfolds is the obligation to release detained persons at the end of “active hostilities,” a point recognized with respect to Guantánamo detentions by a plurality of the Supreme Court in Hamdi v. Rumsfeld. Though the conclusion of a peace agreement is by no means required to conclude that active hostilities have ended, it may well be probative of such an ending, particularly where such agreement calls for severe drawdowns and withdrawal. Even if the relevant armed conflict in Afghanistan (or globally) is found to be ongoing, the question of whether that conflict is the same conflict during which some detainees were captured also deserves a fresh look, as does the acceptability of maintaining detentions based solely on ongoing armed conflicts against Al-Qaeda in other parts of the world (in contrast to a “case-by-case” approach). Further, some countries (for example, Canada) emphasize the expiration of detention authority “as soon as the reasons which necessitated internment cease to exist,” a view not developed in governing Guantánamo-related case law. U.N. human rights bodies, meanwhile, stress the unlawfulness of Guantánamo detentions as arbitrary and indefinite.

In short, limiting possibilities exist, and nearly all begin by taking the facts and the duration of each detention seriously.

Concluding Thoughts

The Agreement with the Taliban leaves unresolved important aspects of how the United States will wind-down its security operations in Afghanistan. As the process is said to unfold in parts, it is critical that the United States pursue clearer terms going forward with all sides — including Afghanistan — that make certain each side’s commitments, and how they are to be enforced.

Image – (L to R) US Special Representative for Afghanistan Reconciliation Zalmay Khalilzad and Taliban co-founder Mullah Abdul Ghani Baradar shake hands after signing a peace agreement during a ceremony in the Qatari capital Doha on February 29, 2020. (Photo by Giuseppe CACACE / AFP) (Photo by GIUSEPPE CACACE/AFP via Getty Images)