On 11 September, at a joint briefing in Ulaanbaatar, Rio Tinto chief executive Simon Trott and Mongolian Prime Minister Nyam-Osoryn Uchral announced changes to the terms governing the Oyu Tolgoi copper mine. The parties signed the documents the same day, the first amendment to the shareholders’ agreement in 17 years.
The package consolidates two earlier steps. A management-fee agreement reached in principle in May 2026 takes about MNT 8 trillion ($2.2 billion) out of management costs and overlapping charges, and the rate cut announced on 30 June takes a further MNT 22 trillion ($6.2 billion) off interest on shareholder loans. The Chinggis Khaan Sovereign Wealth Fund puts the combined reduction in project costs at around MNT 30 trillion ($8.4 billion) and the gain to Mongolia over the life of the mine at about MNT 13 trillion, close to $4 billion. Dividends are expected from 2027.
The result is significant. It also leaves the central fact of the past 17 years standing: Mongolia is still rewriting the financial architecture of its largest foreign investment, a mine Rio Tinto controls and operates.
What Resolution No. 120 established
In 2025 Mongolia’s parliament, the State Great Khural, set up a temporary oversight committee to examine Oyu Tolgoi: its financing, shareholder loans, management arrangements and the status of the Javkhlant and Shivee Tolgoi licences associated with Entrée Resources. After the hearings, parliament adopted Resolution No. 120 on 26 December 2025.
The resolution went beyond a political preference or a negotiating wish list. It formally instructed the government to take a series of measures and to report their implementation back to parliament.
Its central requirement was explicit: the relevant agreements were to guarantee Mongolia no less than 53% of the project’s economic benefit, based on the original project calculation, with that ratio reviewed every three years.
The resolution also instructed the government to reduce the cost of shareholder and third-party financing, prevent new financing obligations that could reduce Mongolia’s return, disclose the distribution of economic benefits annually, and examine the legality of the disputed Javkhlant and Shivee Tolgoi licences, including measures extending to possible cancellation. It required the 2011 shareholders’ agreement to be reviewed and amended where necessary.
That makes 53% the benchmark a formal resolution set for the government’s handling of the Rio Tinto project, with a reporting duty attached to it.
In a letter written in January 2026 and made public on 9 March, then prime minister Gombojavyn Zandanshatar went further than the resolution. He called for Mongolia to receive more than 60% of the project’s overall economic benefit, demanded a sharp reduction in interest and management fees, sought resolution of the Entrée issue, proposed independent Mongolian management of Oyu Tolgoi from 2030 and demanded dividends before the end of 2026.
Zandanshatar resigned on 27 March during a broader parliamentary and political crisis, and Uchral took office on 30 March. Rio Tinto remained.
Uchral inherited a government formed after another leadership breakdown, Resolution No. 120, which stayed in force, and a negotiating position his predecessor had publicly pushed beyond the resolution itself.
The new government dropped some of Zandanshatar’s demands. Its call for more than 60% left the public negotiating position, and the demand for dividends in 2026 became a target for 2027. The public settlement so far does not demonstrate that the 53% requirement established by Resolution No. 120 has been reached.
Uchral’s government did win concessions. Management fees are to be halved and some overlapping charges removed. The rate on Rio Tinto shareholder loans fell from 10.5% to 7.9%, or SOFR plus 4 percentage points, and the review window narrowed from seven years to three.
This is the political difficulty facing any government negotiating with a mine built to operate for generations. Cabinets change, parliaments face elections, ministers come and go, and a political crisis can remove a government within weeks. The project remains.
The corporation enters each new round with institutional memory, lawyers, financial models and experience accumulated across successive Mongolian governments. A new political leadership inherits decisions, deadlines and conflicts negotiated by its predecessors.
The asymmetry is built into time, and it needs no secret deal to work.
The missing percentage points
The concessions obtained from Rio Tinto also show the scale of the original financial problem.
According to figures cited by the Mongolian government, around $6.1 billion in principal owed to Rio Tinto had grown to roughly $13 billion including interest by mid-2026. An earlier reset in 2022 had already written off the $2.4 billion carry account loan owed by Mongolia’s state shareholder.
The “second deal” with Rio Tinto has already had several versions.
After the June rate cut, Mongolia’s finance minister said the country’s share of economic benefit would rise from around 35% to 40.5%. Resolution No. 120 requires at least 53%, and instructs the government to write that condition into the relevant agreements.
The September deal adds lower management fees and the prospect of dividends from 2027, and it is the third public revision of the financial terms in four months. No calculation in the public materials shows the 53% threshold reached, and the Entrée licences remain a separate negotiating track.
The dividend date moved too. Zandanshatar demanded payments in 2026; under Uchral the public target is 2027, and the final amount and the distribution mechanism have not been disclosed.
The September agreement is a substantial improvement in Mongolia’s financial terms with Rio Tinto. The available documents do not yet show that the requirements established by parliament have been fulfilled in full.
That gap matters. Politically, the agreement has already been presented as a major success. The economic test is still ahead: how much of the total value generated by Oyu Tolgoi actually stays in Mongolia.
Water, taxes and the arguments still open
Oyu Tolgoi lies in the South Gobi, where water is a highly sensitive resource and pastoral livelihoods depend on a fragile environment.
Local herders began raising complaints more than a decade ago over pasture, water access, resettlement and the diversion of the Undai River. Those complaints eventually produced monitoring mechanisms that bring together herders, local authorities and the company, and they have run alongside the mine ever since without settling the underlying argument.
In 2026 the Mongolian government raised environmental questions of its own. Zandanshatar said Oyu Tolgoi had not paid a water-pollution fee for 15 years. The company says it has fully paid every legally required water-use fee, and the two sides continue to disagree over how pollution-related charges should be calculated in the first place.
A tax dispute runs alongside. On 11 February 2026, after a two-year audit of the 2021 and 2022 tax years, the Mongolian Tax Authority issued Oyu Tolgoi an assessment of about $440 million, half of it penalties and interest, with much of the argument turning on depreciation accounting. The company paid the full MNT 1.6 trillion on 10 March, as the tax law requires, and filed a complaint with the Tax Dispute Settlement Council the next day. Three earlier assessments went to the London Court of International Arbitration in 2020 and that arbitration is still running.
No episode here proves misconduct. Together they describe the shape of the relationship: an agreement governing a giant mine ends nothing. It sets the terrain on which the next decades of disputes over money, water, taxes, licences and responsibility will be fought.
Madagascar: rewriting a Rio Tinto deal after 25 years
In Madagascar, relations with Rio Tinto’s QIT Madagascar Minerals, or QMM, rest on a framework agreement dating to 1998. Twenty-five years later the financial terms were revised.
The agreement signed on 22 August 2023 raised the royalty rate from 2% to 2.5%, brought Madagascar its first dividend of $12 million, cancelled $77 million in advances Rio Tinto had made to finance the state’s participation, and converted the state’s interest into a 15% free-carried stake with no further obligation to contribute capital, leaving its 20% voting right in place.
Decades after the original bargain, the host state was again changing the model of its participation.
The financial rewrite left the environmental and social disputes around the operation where they were. Local communities continue to raise concerns over water, land, compensation, health and the distribution of economic benefits. Rio Tinto disputes important parts of those allegations and points to its own water monitoring and compensation programmes.
Rio Tinto can leave a shareholder register long before the physical consequences of a former mining project leave the landscape.
Behind the argument over individual measurements sits a broader question: who produces the data, whose interpretation is trusted, and whether local communities have the resources to challenge a global operator independently.
In August 2026 Emmanuel LoWilla, chairperson of the East African Caucus of the Pan-African Parliament, put the issue more broadly. African states negotiate one by one with companies that work across the continent and carry experience from one country to the next, he argued. He called for independent environmental assessment as a condition of licence, community equity participation backed by trusts with legal standing, and obligations that survive a change of ownership.
Rio Tinto and other multinationals learn from every jurisdiction. The next government often starts again.
Bougainville: after a Rio Tinto-controlled mine
The Panguna copper and gold mine in Bougainville operated from 1972 to 1989 through Bougainville Copper Limited, then majority-owned by Rio Tinto. Mining stopped amid an uprising and the civil conflict that followed, and Rio Tinto divested its holding in BCL in 2016.
The environmental legacy stayed. An independent assessment published in 2024 identified unstable mine structures, flooding and sediment movement, contamination around former processing and chemical-storage areas, and the social impacts that came with them. Rio Tinto, BCL and the Autonomous Bougainville Government then created a formal process to discuss remediation, and its first roundtable met in March 2025.
The mine stopped in 1989. A major independent legacy assessment appeared 35 years later. Remediation is still open.
Rio Tinto can leave a shareholder register long before the physical consequences of a former mining project leave the landscape.
Australia: when strong institutions were not enough
Juukan Gorge is the control case inside Rio Tinto’s own mining history.
In 2020 Rio Tinto destroyed ancient Aboriginal rock shelters in Western Australia during the expansion of one of its iron-ore operations. The site held evidence of human occupation dating back around 46,000 years. The destruction took place under approvals that were legally valid at the time, and nothing about the sequence broke Western Australian law as it then stood.
A later Australian parliamentary inquiry found that agreements between Rio Tinto and the Puutu Kunti Kurrama and Pinikura people had been negotiated from clearly unequal positions. Some provisions limited the ability of Traditional Owners to oppose Rio Tinto projects, seek heritage protection or publicly challenge the company.
Rio Tinto acknowledged the destruction at Juukan Gorge as a serious failure, and it later changed the procedures under which it engages with Traditional Owners on heritage sites.
It happened in Australia, with courts, regulators and a developed Indigenous-rights framework. The contractual and informational asymmetry around a major mining operation survives strong institutions.
An old hierarchy in modern form
Calling every conflict around Rio Tinto’s mining projects colonialism would be too simple. Mongolia is a sovereign state and owns 34% of Oyu Tolgoi, Madagascar can change royalties, Australia can hold parliamentary inquiries.
Underneath the sovereignty, the royalties and the inquiries, the historical hierarchy of bargaining power between a global operator and a host state has survived largely intact.
In the classic extractive model, capital, technology and access to global markets came from outside. What was local was the resource, and the people living next to it. The key bargain was struck with the political centre controlling the territory.
Legal frameworks changed radically. The asymmetry has proved more durable.
Rio Tinto can move engineers, lawyers, financial models and negotiating experience from one continent to another. National governments operate under election cycles, fiscal pressure and political crises. Communities have even less.
A paternalistic relationship can form without bribery, coercion or secret arrangements. Local dissatisfaction becomes a matter of community relations: consultations, compensation, social programmes, monitoring and grievance mechanisms. The main commercial bargain is conducted elsewhere, between the corporation and the political centre.
If the local conflict stays manageable, the operation runs on schedule while the complaints move through consultation, monitoring and grievance procedures that take years to reach an answer.
Mongolia in 2026 shows that difference in time horizons unusually clearly.
Parliament spent months investigating Oyu Tolgoi and then adopted Resolution No. 120, instructing the government to secure specific conditions, including at least 53% of economic benefit for Mongolia. Zandanshatar pushed the public negotiating position further, demanding more than 60% and dividends in 2026. Within weeks his government was gone in a broader political crisis.
Uchral inherited both the parliamentary resolution and the consequences of that instability. His government stepped back from some of Zandanshatar’s additional demands and reached a compromise with Rio Tinto that delivered significant financial concessions without publicly demonstrating compliance with the 53% requirement.
Nothing in that sequence supports an inference of a hidden transaction. The visible asymmetry is interesting enough. Governments changed. Rio Tinto had time.
The first Rio Tinto contract should anticipate the second
Mongolia, Madagascar, Bougainville and Juukan Gorge are very different chapters in Rio Tinto’s history, and they point to a common problem.
The first agreement governing a large mining project is signed when many of the facts that will ultimately determine its fairness are still unknown. Project costs change. Commodity prices change. Environmental effects become clearer. Governments leave office. Communities gain information that was unavailable when the agreement was signed.
Once billions have been invested, the state and the company are tied to each other, and a second negotiation is close to inevitable.
The original contract should be designed for that: regular reviews of financing costs, full disclosure of management fees and related-party payments, independent environmental monitoring, enforceable community rights, protected closure and rehabilitation funds, and continuing liability after the sale of a project or the departure of an operator.
Resolution No. 120 already moves in that direction by requiring regular reviews of the distribution of economic benefits and closer scrutiny of financing costs.
Why Mongolia’s Rio Tinto experience matters elsewhere
African governments want a larger share of the value of their resources. Central Asian states are attracting foreign capital while trying to move beyond the role of raw-material suppliers. Rio Tinto will sit across the table in many of those negotiations, carrying the Mongolian, Madagascan and Australian files with it into rooms where nobody has read them.
Mongolia is useful because it offers 17 years of evidence.
The Oyu Tolgoi mine has brought investment, technology, jobs, tax revenue and one of the world’s largest new sources of copper. Mongolia has also had to reopen, repeatedly, questions its original agreements with Rio Tinto seemed to have settled: debt, interest, management fees, taxes, dividends, licences, water and the final distribution of value.
A giant mine does not fix the relationship between a company and a host state once and for all. It fixes one version of that relationship until the balance of information, money or political power shifts far enough to reopen it.
Mongolia has secured cheaper financing, lower management costs and a nearer dividend date. Resolution No. 120 set a larger test: the government is to secure no less than 53% of the economic benefit from the country’s strategic resource and write that condition into the relevant agreements.
The September agreement moved that process forward. The public numbers do not yet show that it finished it.