A tale of two sixes

As COP29 drew to a close in Baku, Article 6 negotiators celebrated. They had finally wrapped up negotiations that had been running since COP21 in Paris in 2015. The rules were set, and the focus could shift to implementation. That meant direct country-country carbon transactions under Article 6.2, as well as a UN-operated marketplace for carbon trading under Article 6.4.

The results one year later are mixed: a tale of two sixes.

Great expectations

A flurry of activity on Article 6.2, with money starting to flow to real-world projects. Singapore completed its first RFP for A6.2 credits and contracted over 2 million credits. It then released a follow-up RFP: this time open to both tech-based and nature-based projects. Norway and Switzerland concluded the first Carbon Dioxide Removal (CDR) transaction under A6.2. It was a small pilot — but real, and legally binding. More of the bilateral deals needed to enable A6.2 transactions were signed: another 30 since Baku, bringing the total to ~100.

Article 6.4, however, has not left the starting line. Some additional foundations were laid: the first approved methodology, several accredited Designated Operational Entities (DOEs, a.k.a. VVBs), and a raft of new guidance notes. But compared to the heady expectations in Baku — and the urgency of the climate crisis — this is glacial progress. There were no projects registered; no credits sold; no carbon dioxide emissions reduced or removed, contrary to expectations set by the Supervisory Body (SB) earlier this year.

Hard times

The progress gap looks set to widen before COP31 in Antalya. The two provisions represent very different philosophies on how best to drive climate action. Article 6.2 is devolved: the buying and selling countries determine how to structure a transaction in compliance with the overarching framework set out in Article 6. Article 6.4 is centralised: the SB runs the show, approving all methodologies, registering all projects, and accrediting all DOEs. This model creates a critical dependency on a single entity, which in turn leads to structural weaknesses in three areas.

Speed

1. The UN is not known for speed, and the Paris Agreement Crediting Mechanism (PACM) reflects this. It took a decade to set the rules. It may end up being several years from those rules being set to the first new projects issuing credits.
2. Only one methodology was approved in 2025 — and this was a simple transition from the Clean Development Mechanism, not a new methodology.
3. Methodologies are unlikely to evolve at pace as the latest scientific evidence changes. That means a methodology approved in 2026 is likely frozen until 2031 or later. This creates integrity risks, as happened in the Clean Development Mechanism (CDM) for renewable energy projects, where the additionality of the projects changed, but the rules did not.

Capacity

1. The SB sets the rules, approves the methodologies, and issues the credits. Yet it needs to do all this with just 12 part-time civil servants. These people have busy day jobs beyond PACM — often leading large chunks of climate policy within their home country. They rotate out of the SB every two to four years due to mandatory term limits (the SB controversially, and unsuccessfully, sought to remove these term limits at COP30).
2. The UN is building a registry for PACM. Not just a central repository of transactions, but a fully-fledged system: account log-ins, transaction flows, and presumably customer support. They will outsource the development work. But the very broad scope and lack of in-house expertise are a recipe for further delays and cost overruns.
3. The SB approves methodologies written by others. Those who reviewed it will likely have rotated out by the time credits are issued. What happens when a question of interpretation arises during a verification event? The real world brings unexpected challenges, and not everything can be covered even within a 100-page methodology. That’s when the deep technical expertise and context that comes from writing the methodologies is crucial — and is missing from PACM.

Funding

1. The SB is supported by a Secretariat of around 60 people, sitting within the broader UN Framework Convention on Climate Change (UNFCCC). This, in theory, provides capacity, continuity, and institutional expertise. This comes at a cost: the annual budget is ~$20m. It is a yearly scramble to secure the necessary funding — a problem for long-term planning.
2. The longer-term funding model is meant to be secured by a levy charged to project developers on credits issued. This, along with the fact that developers are paying the DOEs directly, creates an institutional incentive to issue more credits, which is a risk to credit integrity.
3. 7% of credits issued under PACM are deducted at source. This contribution goes to a good cause, including the UN Adaptation Fund. But the costs will ultimately be borne by buyers. So PACM’s funding relies either on corporate buyers providing significant demand at a premium price — not a safe bet. Or governments doing so, despite the alternative of buying through Article 6.2, at a lower cost and with greater oversight on credit quality and process.

Our mutual friend

PACM is an outlier. Compare it, for example, to CRCF, the EU’s legislation for regulating carbon markets. A6.4 and CRCF share the same objective: issue high-integrity carbon credits. Yet whilst Brussels is famous for its centralising tendencies, A6.4 is much more monolithic than CRCF.

The EU does not seek to do everything in CRCF. It instead regulates the key players within the value chain and leverages their resources and expertise. It approves and oversees Certification Schemes, which issue the credits. It relies on an existing network of National Accreditation Bodies to accredit the Certification Bodies (in UN-speak: DOEs). And its own registry is being designed as a repository consolidating information from across Certification Schemes—rather than one with full account management functionality.

Like financial services, carbon markets are too important to be left unregulated. In finance, central banks write the rules of the game that all players have to follow—but they know their limits. They regulate the banks, which in turn offer direct services to the population, and enforce rules such as money laundering checks. No sane central bank has ever built its own retail banking app. Yet that’s the equivalent task that the Supervisory Body has taken on for carbon markets—it is ill-equipped to do it.

Bleak house

Unless these structural weaknesses in PACM are addressed, one of two bad scenarios could play out over the next couple of years.

• Stasis. PACM could find itself bottlenecked. With hundreds of project developers in the queue, insufficient funding, and weak institutional memory, the mechanism could grind to a halt. This would damage its credibility as developers seek alternative routes to market.
• Rubber stamp. In order to avoid stasis, the SB could take a light-touch approach. This would involve leaving all meaningful scrutiny to DOEs, who have clear financial incentives to allow developers to issue more credits. Then PACM ends up falling victim to the same integrity and overcrediting challenges as the CDM.

Neither outcome is good for the industry, the atmosphere, or the UNFCCC itself. The architecture of A6.4 is next due to be reviewed in 2028. In the meantime, Article 6.2 is shaping up to be the better bet for climate action.

This article is also published on LinkedIn. illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.

Track the real‑world impact behind the sustainability headlines. illuminem’s Data Hub™ offers transparent performance data and climate targets of companies driving the transition.