ANFREL Report: BARMM's 2026 Elections Marked a Milestone, but Intimidation and Irregularities Cloud Integrity

On September 13 and 14, 2026, violence broke out in Cotabato City, the regional capital of the Bangsamoro Autonomous Region in Muslim Mindanao. At least five people were killed and more than 20 were injured. The city is the political center of the contest between the United Bangsamoro Justice Party (UBJP) and the Bangsamoro Federalist Party (BFP). Its violence overshadowed what the Asian Network for Free Elections (ANFREL) and the National Democratic Institute (NDI) call a historic vote.

The September 14 election was the first parliamentary election in BARMM. It came after decades of armed conflict, a negotiated political settlement and repeated postponements. The joint Technical Assessment Mission describes it as a step from "bullets to ballots," in the footsteps of Timor-Leste, Aceh in Indonesia and Nepal. Holding the vote across the region on a single day was itself a major accomplishment. The one exception was some precincts in Maluso, where voting ran into September 15.

The mission visited the region before, during and after election day. On the day itself it observed 102 precincts in every BARMM province, through opening, voting and closing, though clearance protocols and a volatile security situation sometimes restricted its movement. Outside Cotabato City it found only sporadic reports of violence, and the Commission on Elections (COMELEC) called the day "generally peaceful." The mission concluded that conditions still fell short of letting voters, electoral boards, poll watchers and observers take part without fear. Controlled voting and intimidation limited free participation even where no violence was formally reported. It pointed to an unfinished normalization process, widely available firearms, ongoing clan feuds, and a contest between two parties both rooted in the Moro Islamic Liberation Front. Security forces were widely seen as necessary in high-risk areas, but some were stationed close enough to voting centers to intimidate voters.

The mission treats the official turnout of 84.34 percent with caution. Some 317 precincts reported 100 percent turnout, and it found large discrepancies in granular results. It asks COMELEC and law enforcement agencies to investigate outlier precincts and municipalities and to disclose any irregularities they find.

Electoral administration was in line with the 2022 and 2025 national and local elections that ANFREL previously observed. The automated election system again gave the process a strong foundation, with incremental improvements to ballot papers and data transmission. Results transmission and canvassing were quick and transparent, which likely eased tensions between rival parties. Observers still recorded late delivery of materials that cut into early voting hours, skipped indelible ink checks and non-voters feeding ballots into counting machines. Poll watchers frequently marked ballots for voters or assisted them, both election offenses. Electoral board members, outnumbered by poll watchers, may have been unable or unwilling to enforce the rules. The mission also recorded vote buying in the open inside voting centers, voters bussed in by parties, and illegal campaigning and local officials near polling places. It saw pre-shaded ballots in three precincts, showing the practice went beyond Maluso, Basilan, where COMELEC found and replaced more than 16,000 of them.

The campaign itself was peaceful and vibrant in the events observed, but the playing field was uneven. Stakeholders repeatedly flagged incumbents' access to state resources, including block-grant programs, and large gaps in money, party machinery and experience among the parties. Many believed those gaps decided outcomes. A revised system for most sectoral seats replaced selection by sectoral assembly with a region-wide contest. That raised campaign costs and, in most cases, disenfranchised grassroots sectoral organizations. Online campaigning relied mainly on Facebook and made heavy use of AI-generated videos, jingles and images. Manipulated images that exaggerated crowd sizes misled voters. COMELEC's Committee on Kontra Bigay, aimed at vote buying and abuse of state resources, and its Task Force Katotohanan, Katapatan at Katarungan, aimed at disinformation, showed institutional commitment, but the mission says their effectiveness needs further assessment.

Voter education was substantial, online and offline. Ballot pictures and logos helped, and the election debate was a positive step. Repeated postponements gave more time for outreach but forced frequent revisions to materials, straining civil society. Given limited familiarity with parliamentary government, linguistic diversity, uneven literacy and patchy connectivity, many stakeholders thought the reach insufficient, leaving isolated communities more open to undue influence. Undervotes in the sectoral contest reached 23.46 percent, and overvotes in the party contest 3.95 percent.

Legal changes in the two years before the vote weakened certainty about the rules. Women held a 30 percent quota among party nominees but won only 13 of 80 seats, about 16 percent, without an enforceable placement rule such as a zipper system. The mission's calls now fall to COMELEC and the police: investigate the outlier precincts and make the findings public.