Crossing the Rubicon: Are Myanmar’s ethnic armies prepared to go all in?

After more than a year of military victories, ethnic armed organisations (EAOs) are in control of large portions of Myanmar’s peripheries. These advances, however, have not brought greater stability to the borderlands. Still entrenched in the centre, the relentlessly stubborn regime has adopted a strategy to deny any return to normality through enduring blockades and an ever-escalating air campaign. Meanwhile, neighbouring countries remain reluctant to recognise emergent authorities in the absence of a negotiated settlement between them and the regime, reducing the viability of a cross-border development model or large-scale humanitarian relief.

Mostly unwilling to negotiate with the regime, powerful opposition actors in the north and west appear poised to take the fight to the centre of the country, raising the spectre of both regime and state collapse. But not all forces are capable of expanding the war. Political and military campaigns by more democratically minded forces in the southeast have stagnated, a trend set to continue following the United States’ recent withdrawal of assistance. With the country fundamentally transformed by four years of devastating violence, it is difficult to predict what will come next.

Routed in Rakhine

Following a six-month siege, fighters from the Arakan Army (AA) overran the Border Guard Police No. 5 base in Maungdaw on 8 December 2024, stripping the regime of its last position along the Myanmar–Bangladesh border. Twelve days later, on 20 December, the AA captured the Regional Military Command (RMC) in Ann Township. The sprawling RMC lies along the main road linking northern Rakhine to inland Myanmar and was a critical pillar of the regime’s military presence in the state. Continuing its thrust southward, the AA was in control of 14 of the state’s 17 townships by New Year’s Day. At the time of writing, in February 2025, only Rakhine’s capital of Sittwe, the important port area of Kyaukphyu and a remote island, Manaung, remain in regime hands.

The outcome of these pivotal battles provides several important indicators of the regime’s overall strength and capacity to wage war. Most notable was its failure to reinforce the RMC at Ann. Though the army had struggled to launch large-scale counter-offensives since late 2023, the response pattern to recent attacks in Rakhine has been even more limited. This failure raises questions about the regime’s handling of its conscription drive, first announced in February 2024. According to open-source reports, training of the tenth batch of conscripts began in February 2025, meaning that the regime has probably gathered at least 35,000 new recruits and perhaps as many as 50,000.

To wage counter-offensives the army must use its new recruits to rebuild its heavily depleted Light Infantry Divisions (LIDs) and Military Operations Commands (MOCs), which are both referred to colloquially as mobile strike brigades. These units are traditionally assigned to spearhead major operations or counter EAO offensives. However, it is not clear how the regime is using its conscripts. One possibility is that initial cohorts of recruits were deployed to backfill depleted local units across the country in an ad hoc fashion, delaying the reformation of strike brigades. 1 Another possibility is that newly rebuilt strike brigades are undergoing extended training and therefore are not ready for combat, which could explain the failure to reinforce the RMC at Ann.

On 29 January the regime reportedly deployed a large column of soldiers from the 99th LID to halt the AA’s westward march along the border of Rakhine, where the unit immediately suffered losses. The size of the contingent was potentially six or seven battalions, which would indicate that the 99th has been largely rebuilt, but the late deployment suggests it was not ready in December. 2 Rushing conscripts into battle will reduce their combat effectiveness.

It also seems the regime is running into increasing difficulty with its conscription effort. Initially, many conscripts were offered inducements of various kinds while some volunteered at the behest of the Union Solidarity and Development Party (USDP), the military’s proxy party. Once the pool of these potential recruits was depleted, conscription became increasingly coercive. Then, on 24 January, the regime finally promulgated the People’s Military Service Rules. Prior to this, conscription practices had essentially been unofficial, with significant loopholes such as a potential conscript being able to pay someone to replace them. While the regime cannot ensure that conscription is enforced uniformly across the country, promulgation of the rules has coincided with an effort to crack down on draft evasion. In January, for example, the regime tightened exit restrictions for conscription-aged individuals seeking to leave the country.

The regime’s handling of conscription will have implications for the trajectory of the war. An ability to rebuild and refit its strike brigades should in theory allow the regime to launch counter-offensives and reinforce critical areas such as Sittwe, which the AA moved to besiege in early February. Whether the regime can send significant reinforcements to the city will be an indication of its ability to properly utilise conscripts in a timely manner. The events of the last year show that isolated regime bases and encircled cities eventually fall if they do not receive adequate reinforcements, as long as EAOs are willing to accept high casualties and severe humanitarian consequences over an extended timeline.

A new problem

Though it has conquered most of Rakhine, there is no obvious way for the AA to end the war soon and redirect its energy towards reconstruction or economic development. The group’s predicament is best illustrated by the electricity situation in the state. Prior to the coup, most of Rakhine’s power was carried over the mountains from Magway, but the regime shut off electricity once hostilities escalated and the grid became overburdened. At present, the largest operable source of local electricity is a Chinese-built gas plant at Kyaukphyu, where the China–Myanmar pipeline begins. Even if the AA seizes the power plant, it cannot operate it without Beijing’s participation, or provide power to the entire state.

In addition to suffering from a blanket electricity outage, Rakhine remains under blockade by the regime, and neither Bangladesh nor India appears ready to normalise relations with the AA or open cross-border trade and investment at a scale that could solve the state’s problems. As a result, Rakhine is suffering from severe shortages in essential goods and two million people are at risk of famine. The regime has also intentionally created a cash shortage in the state, making it hard for the AA and the civilian population to access goods and services even when they are available. There is no way for the AA to provide reliable banking services, and a blanket internet shut-off, which the AA also cannot remedy, makes mobile alternatives less viable. Rakhine people living in other parts of the country are facing renewed racial discrimination instigated by the regime and cannot easily access passports, and there is no obvious way for the AA or other non-state actors to address this sort of problem.

The AA also has no way to counter the regime’s surveillance uninhabited aerial vehicles or its airstrikes, which continue unabated. On 9 January, for example, regime aircraft struck a suspected fuel-storage depot in the Rohingya village of Kyauk Ni Maw, leading to a large explosion and fire that destroyed the village and killed scores of civilians. Two days later, a regime airstrike on the central market in Kyauktaw killed numerous civilians along with a Rakhine political leader, U Waing Maung. The pattern appears to be one of intelligence-driven strikes on AA leaders and assets located in civilian areas, which the military has no qualms about targeting. The monthly number of airstrikes continues to reach new highs, and the regime generally appears willing and able to bomb its opponents indefinitely.

Security dilemmas

On 20 January, Beijing revealed that it had brokered a formal written ceasefire between the junta and the Myanmar National Democratic Alliance Army (MNDAA). Following the agreement, China re-opened cross-border trade and services to the MNDAA’s territory. In doing so, China has demonstrated its willingness to confer recognition upon borderland groups if they agree to co-exist with the central regime.

According to various reports, the deal involved an MNDAA agreement to gradually withdraw from Lashio, which it captured in August 2024. Should that happen, the MNDAA’s withdrawal would probably be followed by a hybrid arrangement in the city, rather than an outright return to regime control. For example, the MNDAA could allow regime administrators to return while maintaining its own police force. Alternatively, a third party could provide security or administrative services and be responsible for disbursing shared tax revenues between the MNDAA and junta. Hybrid arrangements such as these have long existed in Myanmar.

Following the agreement, talks between the regime and Ta’ang National Liberation Army (TNLA) began in Kunming, China, on 15 February, and China reportedly plans to host sequential negotiations involving the Kachin Independence Army (KIA) followed by the AA. But brokering deals between the regime and these groups is less straightforward than with the MNDAA. Firstly, China has much greater influence over the MNDAA than it does over any of the other groups actively fighting the regime. Also, the MNDAA’s area, Kokang, is far from central Myanmar and an inability to control it does not pose an existential problem for the regime. Another factor is that the regime had already agreed in principle to recognise the MNDAA’s authority over Kokang last year, whereas it has not hinted at any similar concession for other opponents.

But a more fundamental issue is the security dilemma that the TNLA, KIA and AA face in relation to the junta, and vice versa. To be secure in their new enclaves, the EAOs must establish buffer zones, which means controlling parts of central Myanmar either directly or through proxy forces. Likewise, the regime must also ensure control over a buffer space between core territories and EAO strongholds. This makes it exceedingly difficult for either side to agree to a ceasefire — let alone a long-term peace agreement — based on the current lines of control. It is highly unlikely, for example, that the regime could tolerate a permanent TNLA presence on the edge of Mandalay or an AA presence on the edge of Magway, where its weapons factories are located. Any ceasefire in these areas is likely to be only tactical unless it is somehow enforced by China.

In the absence of a deal with the centre, EAOs will continue to face blockades and relentless air attacks designed to cripple nascent administrations and demoralise civilian populations. Under the status quo, there is virtually no way for EAOs to provide electricity, internet, banking or passports at scale without cooperation from the central regime, which continues to deliberately deny these services as part of its counter-insurgency strategy. While Myanmar’s neighbours could probably cooperate with EAOs to provide partial solutions, they may not formalise relations with these actors or make large-scale investments in their territories without acquiescence from the central government, and this cannot happen without a wider deal involving the regime.

This reality leaves the EAOs in a serious predicament. To achieve stability via a negotiated agreement, groups such as the AA, TNLA and KIA would potentially need to cede some of their captured territory to the regime, as the MNDAA has purportedly agreed to do. Alternatively, serious negotiations could come about if the regime were able to forcefully recapture territory, but as explained above, it is not clear when, or indeed if, it will be able to gather the necessary strength to do so.

Furthermore, it cannot be assumed that additional battlefield setbacks or increased pressure on the centre will force the regime into negotiations. The Myanmar military is an institution driven primarily by ideology rather than material interest. EAO control over portions of central Myanmar would run counter to the military’s fundamental vision of itself as the protector of the Bamar nation state. Moreover, the military is likely to interpret any EAO presence in central Myanmar as an existential threat, irrespective of any agreement. From this perspective the logical course of action is to push the EAOs back rather than allow them to establish a foothold in central areas under the cover of ceasefire.

Taken together, these dynamics are driving an increasingly binary choice for opposition forces that wish to rule over their newly captured territories. Unless they are willing to forfeit portions of their gains, EAOs seeking stability at home may feel they have no choice but to march into central Myanmar and attempt to permanently expel the military.

A power vacuum looms

In a sign of which path it may choose, the AA has already begun probing attacks into the neighbouring regions of both Ayeyarwaddy and Magway. Though it appears that the AA will prioritise gaining control over Sittwe (and maybe Kyaukphyu) first, it could increase support for proxy forces operating in adjacent areas in the meantime. For its part, the regime will probably attempt to avoid major conflict in Shan and refocus its attention towards preventing the AA’s breakout into Ayeyarwaddy and Magway. Should it take place, however, a major AA offensive into Ayeyarwaddy Region would directly threaten Yangon and lead to mass displacement in an area of the country unfamiliar with conflict.

Similarly, the KIA is expected to make another push into northern Sagaing Region, where it has supported and fought alongside the Kachin-People’s Defense Force (PDF) since the war began. Though it has not achieved territorial gains as swiftly as other northern EAOs, the KIA continues to make steady progress and may soon capture the strategic crossroads at Bhamo, which guards the critical river and road links from Kachin to Sagaing, Mandalay and northern Shan State. The KIA has also proved largely immune, so far, to Chinese demands to halt its offensives. 3

Sagaing is also witnessing intensifying competition between the Kachin-PDF and the Mandalay- People’s Defense Force, which operates under the command of the TNLA. The KIA and TNLA are embroiled in an ongoing territorial dispute in northern Shan, and this translates into contests over buffer areas where their respective PDF allies operate. To complicate matters further, several non-aligned Bamar armies continue to gain power and influence in central Myanmar and are seeking to position themselves ahead of the National Unity Government (NUG). This dynamic offers an opportunity for the regime, which can drive factionalism by negotiating with certain PDFs but not others. China has also capitalised on the fragmented landscape by funnelling arms and cash to Bamar forces that are not aligned with the NUG, according to various sources.

Another unknown is how the TNLA will respond to China’s push for a deal with the regime. On one hand, the MNDAA’s ceasefire with the regime isolates the TNLA from its strongest partner, thereby limiting its military options. On the other, it incentivises the group to step up its cooperation with various PDFs, which are less likely to seek de-escalation in the short term. The first round of talks between the TNLA and the regime ended on 19 February without any agreement.

Although ground fighting between the two had largely subsided by late 2024, the regime continued to fly airstrikes against TNLA military and administrative targets ahead of the talks in Kunming. Though it cannot afford an escalation in Shan, the military has a long history of pressuring the TNLA ahead of negotiations, a tactic that has backfired on multiple occasions by triggering a TNLA response. A resumption of major hostilities along the Mandalay–Shan border, where the TNLA has entrenched itself, could pull the MNDAA back into the fight, put pressure on Mandalay City and pose an existential threat to the regime.

While threats to the regime’s survival are clearly growing, they derive primarily from northern EAOs that remain largely aloof from the NUG and its key ethnic allies, the Karen National Union (KNU) and a coalition of Karenni ethnic forces. Over the last year, the Karenni have suffered reversals against the junta, while the KNU’s progress has stalled along the Myanmar–Thailand border, leading to a stalemate that persists as of early February 2025. These setbacks are largely attributable to disruptions in the supply of arms and ammunition, as Thailand has significantly cracked down on the border market, and the United Wa State Army, the largest supplier of arms to opposition groups, has halted sales at the behest of China. The problems faced by the Karenni and the KNU are likely to be compounded by cuts to United States Agency for International Development funding, as both groups rely heavily on US technical, financial, humanitarian and political assistance, as does the NUG.

The balance of power among opposition forces will matter greatly for any post-regime scenario. The NUG and its allies propose democratic federalism as a solution to Myanmar’s lasting conflict, while groups like the TNLA, AA and MNDAA show less interest in federalism or broad coalition-building and are overtly authoritarian. With the former camp growing relatively less influential, the chances are decreasing that any actor or coalition will be able to step in to fill the vacuum at the centre and deliver democratic reforms should the junta collapse. Instead, the current trend of incremental regime decline is coinciding with a rise in factionalism, ethno-national military autocracy and state failure.