Across the EU, Financial Market Participants (FMPs) face an increasingly urgent compliance challenge, one that sits not in the complexity of the rules themselves, but in the absence of agreement on how to apply them.
The regulatory texts that require global norms screening (SFDR’s Principal Adverse Impact indicators, the ESMA Naming Rules, Paris-Aligned and Climate Transition Benchmarks, and a growing set of national fund labels) all reference the same underlying frameworks: violations of the UN Global Compact (UNGC) Principles and the OECD Guidelines for Multinational Enterprises (MNE). Yet, in the absence of formal guidance from the European Commission, ESMA, or its members on how to interpret those frameworks, supervised entities are navigating a patchwork of expectations that varies not just between jurisdictions, but sometimes within the same regulatory body. The Commission’s November 2025 SFDR 2.0 proposal makes this more urgent still. Global norms compliance is set to play a central role in the new fund categorisation framework, which means the absence of formal guidance on how to identify and apply these screens is no longer a background concern. For many FMPs, it is already shaping day-to-day compliance decisions.
Why supervisors disagree on global norms screening
It is worth noting at the outset that the authors of the UNGC Principles and OECD MNE Guidelines never intended for them to be used as investment screening tools (a challenge that sits at the root of the broader problem). They cover a wide range of corporate conduct: human rights, labour standards, environmental responsibility, anti-corruption, competition, taxation, and consumer interests. A strict reading treats all chapters as equal.
In practice, however, that equivalence has not held. Certain supervisory authorities, through informal guidance and supervisory dialogue, have signalled that human rights and environmental violations carry the most weight. Competition, taxation, and bribery/corruption issues may attract comparatively less supervisory attention.
There is a coherent argument behind this. Environmental damage and human rights abuses cause harm that is direct, lasting, and often irreversible, affecting communities, ecosystems, and future generations across borders. By contrast, competition and taxation infractions, however serious, tend to play out within specific economic and legal contexts and their harm is more commonly financial in nature. A €405 million GDPR fine and a toxic tailings spill contaminating 100 km² of waterways are arguably both regulatory violations, but they are not the same order of harm.
The difficulty is that this prioritisation has never been formally written into EU law or confirmed as a common supervisory approach through ESMA. Through direct engagement with regulatory bodies across the EU, Clarity AI has found that supervisory expectations are not uniform. Some authorities apply a strict, equal-treatment reading; others expect supervised entities to focus first on the most grievous breaches. This divergence runs not just between member states, but even between different parts of the same supervisory organisations.
Comprehensive vs. targeted global norms screening: two approaches explained
The frameworks themselves do not prescribe a hierarchy, and, as a general rule, we do not believe one should be assumed, particularly by FMPs seeking to minimise supervisory and enforcement risk. Competition, for instance, has its own dedicated chapter in the OECD MNE Guidelines (Chapter X), yet certain regulators treat it as a lower-order concern than human rights or environmental violations.
That gap between what the frameworks say and how they are being applied in practice is precisely what FMPs are left to navigate. Not only does this heap regulatory uncertainty into the investment industry, it also risks confusing end investors who may not understand why particular companies are excluded by one asset manager but not another.
For that reason, and following extensive dialogue with both market participants and regulatory authorities, Clarity AI has developed two distinct profiles for UNGC and OECD MNE screening that reflect these two positions:
| Comprehensive Profile | Targeted Profile | |
|---|---|---|
| Scope | All UNGC Principles and OECD MNE Guidelines chapters — competition, taxation, bribery/corruption, consumer interests, human rights, environment | Human rights and environmental chapters prioritised; competition, taxation, and corruption issues included but weighted lower |
| Hierarchy | No chapter treated as more important than another | Reflects de facto supervisory emphasis observed in some EU jurisdictions |
| Suited for | Firms seeking full letter-of-the-law coverage and maximum breadth. Better where supervisory expectations are unclear to be sure of minimising compliance risk. | Firms in jurisdictions where supervisors have signalled a grievous-breach focus. May be riskier where supervisors have not made it clear they recognize an implicit hierarchy of breaches. |
| Example of violation under both profiles: Major environmental disaster | Flagged Toxic water pollution from tailings disposal via a major mining company joint venture (UNGC 7–9; OECD Ch. VI) |
Flagged Toxic water pollution from tailings disposal via a major mining company joint venture (UNGC 7–9; OECD Ch. VI). Environmental damage meets elevated threshold under targeted approach |
| Example of violation under the comprehensive profile only: Data privacy breach | Flagged €405M GDPR fine to Global technology platform for children’s privacy violations (OECD Ch. VIII) |
Not flagged Data privacy/consumer interests breach does not meet prioritised threshold |
| Source: Clarity AI | ||
These profiles are not competing interpretations of the same methodology. They reflect two defensible positions in an environment where, despite operating within a single market, supervisory expectations genuinely differ. For firms operating across jurisdictions, running both profiles in parallel, and examining the gap between them, is itself a useful analytical exercise: the companies that appear in one screen but not the other are precisely those where the interpretive question matters most.
How screening approach affects portfolio exposure: what the data shows
The difference between the two profiles is not marginal. It reshapes the compliance picture substantially by market cap, by sector, and by type of corporate exposure.
Large caps move out of focus; smaller companies move in
Under the comprehensive profile, large-cap companies (above USD 10 billion in market capitalisation) account for 56% of all flagged companies, with their violations concentrated in product safety, data privacy, anticompetitive practices, and damage to biodiversity, spanning both the OECD consumer interests chapter and OECD and UNGC environmental principles. Under the targeted profile, that share falls sharply to 39%.
The companies that move in to fill that gap are small and mid-caps, rising from 44% of flagged companies under the comprehensive profile to 61% under the targeted profile. Violations of small and mid-caps are different in character: they are concentrated in pollution, toxic emissions, waste, and health and safety failures, exactly the categories the targeted screen is designed to focus on (n.b. the same breaches are also identified under the comprehensive profile, which offers a comprehensive view of violations).
Mega-cap companies (above USD 200 billion) sit at the far end of this dynamic. Under the comprehensive profile, two-thirds of their violations are linked to data privacy and security: serious breaches of OECD consumer interest guidelines that have attracted significant regulatory enforcement under GDPR and equivalent frameworks. Under the targeted profile, they are absent entirely: not necessarily because they are seen as compliant by supervisors, but because the nature of their violations places them below the grievous-harm threshold.
The implication for portfolio management is direct. A fund with significant exposure to large or mega-cap technology companies will show a materially different violation profile depending on which screen is applied, and that difference has consequences for fund labelling, PAI reporting, and the defensibility of exclusion decisions. Firms that have never examined that gap may be operating with an incomplete picture of their own exposure.
Sector exposure concentrates in heavy industry
The sector-level picture follows a similar logic. Under the targeted profile, Metals & Mining and Oil, Gas & Consumable Fuels together account for 44% of all flagged companies, compared to 28% under the comprehensive profile. These sectors generate the majority of environmental damage and human rights cases, exactly the violations the targeted screen is designed to surface (n.b. these violations are also flagged in the comprehensive approach, but they account for a relatively smaller number of the total violations).
Media & Entertainment and Software & Services sit at the opposite extreme. Combined, they represent 12.5% of violations under the comprehensive profile. Under the targeted profile, they account for none. Their violations (data privacy enforcement, antitrust cases, misleading advertising) are real and sometimes financially significant, but they do not meet a grievous-harm test.
For portfolio managers and risk teams, the screen choice has direct consequences for sector-level regulatory compliance and reporting. A technology-heavy portfolio may look substantially cleaner under a targeted screen. Such an approach may be justified if the market participant is confident their supervisor will adopt a pragmatic approach and prioritise human rights and environmental issues. However, it could carry risk if the supervisor applies a literal interpretation of the guidelines. For funds with ESG naming obligations or sustainability-linked mandates, this distinction matters because the interpretation by ESMA supervisors may not be aligned across the entire EU.
What SFDR 2.0 means for global norms screening
The fragmentation we describe here is a product of regulatory incompleteness, not regulatory intent. SFDR 2.0 may begin to close the gap. The Commission’s November 2025 proposal gives global norms a central role in fund categorisation, featuring as it does as criteria (c) in both the PAB and CTB exclusions, which are required across the three sustainability labels.
The final shape of SFDR 2.0 remains uncertain. However, assuming the PAB and CTB exclusions are retained, we believe the Commission, either directly or through delegation to ESMA, should issue some guidance on supervising OECD and UNGC breaches in a harmonized fashion across the EU. This will help ensure that FMPs can adopt an aligned approach without fear of enforcement.
Until then, the supervisory landscape will remain uneven. Different regulators have different interpretations, often within the same market. For FMPs, the question is not which screen is objectively correct. It is which approach is defensible in their specific regulatory context, properly documented, and consistent with the expectations of the supervisors they answer to.
In this environment, methodological transparency is the baseline, not a nice-to-have. Firms that can demonstrate they understood the interpretive choice they made are in a fundamentally stronger position than those that defaulted to a single approach without examining the implications.
Author Information
