The free Evidence Brief

One useful company story, once a week.

The week's most useful company-conduct finding, with the sources. No account. No card. Unsubscribe any time.

← All research

The Worst Companies for the Environment in 2026 (Scored)

environmentfossil fuelsmining
Mashinii Research ·

ExxonMobil, Chevron, Alliance Resource Partners and PBF Energy each score −80 on Mashinii's Planet-Friendly Business dimension—the lowest result among the six companies examined here. Rio Tinto and Berkshire Hathaway's Class B record are both −50, correcting the earlier −80 and −90 figures.

These are selected examples of companies with poor environmental scores, not a comprehensive ranking of every business in the world. “Worst for the environment” also depends on what you measure: greenhouse-gas emissions, local pollution, water damage and ecosystem loss cannot be reduced to a single pollution-volume league table.

Which companies score worst in this comparison?

The table uses the latest stored environmental assessment for each linked record, dated 26–27 April 2026 and checked on 8 October 2026. It does not imply a new October assessment of every operation. Equal scores are ties.

Company recordBusiness contextPlanet-Friendly Business
ExxonMobilOil, gas and petrochemicals−80
ChevronOil, gas and refining−80
Alliance Resource PartnersCoal operations and royalty interests−80
PBF EnergyPetroleum refining−80
Rio TintoMetals and mining−50
Berkshire Hathaway, Class B recordDiversified holding company−50

Higher is better on the −100 to +100 scale. A −80 score is an assessment, not an emissions quantity: it does not mean a company emits 80% more than another. A zero elsewhere in the database is not proof of no environmental harm, and an unscored company cannot be assumed to be clean.

The company links let you inspect the evidence rather than rely on the headline. See how Mashinii works for the wider approach, and compare environmental performance separately from worker treatment, community impacts and other values.

ExxonMobil and Chevron: distinguish pollution cases from climate totals

ExxonMobil and Chevron tie at −80. Their records merit scrutiny, but an enforcement case involving certain facilities does not itself measure the entire company's global carbon footprint.

The EPA's October 2017 ExxonMobil settlement announcement concerns Clean Air Act allegations at eight petrochemical facilities in Texas and Louisiana. It describes requirements for flare monitoring, controls and other measures to reduce air pollution. This is a dated record of allegations resolved through a settlement, not a new 2026 finding against every ExxonMobil site.

The EPA's October 2018 Chevron settlement summary addresses claims concerning accidental-release prevention and refinery safety, including improvements across its then-domestic refinery network. It also covers delayed release reporting. Those facts support scrutiny of operational conduct; they should not be described as a direct calculation of lifecycle greenhouse emissions.

For a closer comparison of climate-specific evidence, read Shell vs BP vs ExxonMobil. Keep the reporting year, emissions boundary and legal status visible when comparing claims.

PBF Energy: the local pollution record matters too

PBF Energy's −80 record shows why a discussion about the environment should not stop at carbon. The City of Martinez identifies Martinez Refining Company as PBF-owned.

In February 2026, the Bay Area Air District announced a $10 million penalty, compliance improvements and another $600,000 for environmental projects following a joint prosecution concerning the refinery. The authority says the judgment addressed violations from early 2020 through late 2024, including the 2022 spent-catalyst release. It explicitly excludes the February 2025 fire, which was being addressed separately.

That distinction prevents both understating the remedy and wrongly claiming one settlement closed every incident. Readers living near a facility need site-level information; a parent-company score is only a starting point.

Alliance Resource Partners: more than a coal-only label

Alliance Resource Partners scores −80. Its 2025 annual filing identifies coal operations alongside oil-and-gas and coal royalty segments. Describing its exposure simply as a single coal mine would miss the broader business.

However, the score should not be presented as an automatic mathematical penalty for a sector label. To assess a specific company's footprint, inspect the operating evidence and the scope of its assets. Production, the subsequent use of a product and the remediation of an operating site are related but different environmental questions.

Rio Tinto: corrected score and the importance of Scope 3

Rio Tinto's linked record is −50, not the previous −80. It therefore does not share the lowest score in this selected group. Different environmental questions still remain relevant to mining, including emissions, water, waste and site closure.

Rio Tinto's own 2025 climate reporting reports 575.7 million tonnes of CO₂-equivalent Scope 3 emissions on an equity basis, approximately 18 times its Scope 1 and 2 total. This is an attributed company disclosure, not a Mashinii estimate. It illustrates why an operational-emissions claim needs a clear boundary: value-chain emissions can be much larger.

Do not combine emissions figures with different boundaries or add companies' value-chain totals without considering overlap. Our mining environmental comparison offers a sector-specific starting point for inspecting individual records.

Berkshire Hathaway: withdraw the unsupported “worst” claim

The previous article called Berkshire Hathaway the lowest-scoring company at −90. That is not the latest value for the linked Class B record, which is −50. The separate Class A record currently shows −40. Those inconsistent record-level assessments should not be interpreted as different environmental conduct caused by owning a different share class.

We therefore do not claim Berkshire is worse than the oil majors or manufacture a combined company score. The table identifies the exact record used, and the share-class discrepancy is a limitation needing reconciliation. Nor does the existence of a holding-company score prove that every subsidiary and minority investment has been consolidated into a complete carbon inventory.

How to use a list of environmentally harmful companies

Before drawing a conclusion, ask three things:

  1. What is being measured? A conduct score, tonnes of emissions and local pollution incidents answer different questions.
  2. What is the boundary and date? Check the facility, subsidiary or parent, the reporting year, and whether a source covers operations or the value chain.
  3. What does the source establish? A complaint alleges, a settlement resolves specified claims, and a company disclosure needs attribution and scrutiny. A recent webpage update does not make an old case a new event.

The four −80 records are the poorest results in this comparison. That is a useful reason to examine their evidence, not proof that they are the world's four largest polluters. Search the company database to inspect the businesses relevant to your own decisions and follow the original sources.

Reviewed and corrected 8 October 2026: preserved the article's search focus, corrected Rio Tinto and Berkshire figures, disclosed share-class inconsistency, clarified the selected sample, added the Martinez enforcement record and removed unsupported global-ranking and portfolio-exposure claims. Original publication date preserved.

Privacy policy