Investment Screening Between Discretion and Arbitrariness – Charting the Borderlands
Authors: Steffen Hindelang (CELIS Institute; Uppsala University), Helene Schramm (CELIS Institute) and Liam McGrath (CELIS Institute)
Introduction
Foreign investment screening regimes are built on discretion. Screening decisions often turn on assessments of open-ended lists of risks to ‘national security’ – an inherently ambiguous concept (J Benton Heath 2022, p. 295), which is constantly evolving as States adjust to new threats and vulnerabilities (OECD, 2022, p. 16). The resulting broad margin of appreciation for screening authorities is justifiable, given their task to perform inherently predictive assessments under a high degree of uncertainty. Yet discretion, however wide, is not unlimited. Prima facie, international investment law, as reflected in international investment agreements (“IIAs”), provides an outer boundary to the exercise of national security discretion. In particular, it proscribes arbitrary conduct. Yet drawing the line between a lawful exercise of discretion in matters of national security and conduct that is arbitrary requires arbitral tribunals to mediate between two inherently open-ended concepts.
Tribunals could easily find themselves between Scylla and Charybdis. On the one hand, they risk being perceived as second-guessing highly political decisions that touch upon the core of State sovereignty. On the other hand, excessive deference may reduce investment protection to an empty promise in an age of ever-expanding securitisation.
This exercise can succeed only if the interests at stake are carefully identified and weighed. The greater the uncertainty surrounding the measure, the more compelling the public interests invoked must be; and the factual basis underpinning those public interests must not merely be asserted, but substantiated. Put differently, national security discretion is not a carte blanche: while States may enjoy a margin of appreciation, that margin must be anchored in demonstrable facts and broadly proportionate reasoning.
The difficulty of this exercise is illustrated by the United Kingdom’s 2022 decision concerning Newport Wafer Fab.
I. Substantive Review Standards under Investment Screening Frameworks
At the core of a given FDI screening process lies the screening authority’s substantive review of the investment at hand, the outcome of which determines whether the investment is cleared, cleared with commitments, or blocked (or unwound). While the specific wording varies across jurisdictions, the typical standard of review centres on whether the investment presents a risk to national security (e.g., in the United States, United Kingdom) or closely related terms, such as public order and security or essential security interests (e.g., in Germany, France). A shared feature across many screening regimes is the absence of any extensive definition or closer guidance as to how the concept of national security should be interpreted. As a result, screening authorities are largely left to their own devices in elaborating what constitutes a threat to national security in any given case, which affords them significant discretion (Gemmi, 2025, pp. 161-162).
Investment screening regimes have been described as a ‘regulatory black box’ (Eichensehr and Hwang, 2023, p. 552; Bian, 2021, p. 568). The predictive and often confidential nature of screening makes it difficult to distinguish between decisions grounded in genuine security concerns and those influenced by industrial, political, or geopolitical considerations (Kokkoris, 2024, pp. 156-158). Domestic judicial review of screening decisions may offer limited comfort, given that courts in many jurisdictions tend to defer to the executive authorities’ assessments in matters of national security (see, e.g., De Kok, 2026; Mercado, 2020) or do not review the substantive aspects of a decision at all.
The United Kingdom’s 2022 Newport Wafer Fab decision, which is discussed next, illustrates how an open-ended national security standard can play out in practice, while also raising questions about the rationality, consistency, and proportionality of the screening authority’s exercise of discretion.
II. Newport Wafer Fab: A Case Study
On 16 November 2022, the UK’s Secretary of State for Business, Energy and Industrial Strategy (“the Secretary of State”) made a Final Order (“Order”) pursuant to the National Security and Investment Act 2021 (“NSIA”), requiring Chinese-owned Nexperia BV (“Nexperia”) to divest at least 86% of its shares in Newport Wafer Fab (“NWF”), one of the UK’s largest semiconductor plants. Nexperia had previously completed its acquisition of 100% of the shares in NWF in July 2021 (Arnold & Porter, 2022), and the government intervention was enabled by the then newly established retrospective call-in powers under the NSIA, which had entered into force on 4 January 2022. Four features of the case are significant.
First, there was no consensus within the UK government as to whether the acquisition presented a security risk (Morrison Foerster, 2022). While the government had the possibility of intervening in the transaction at the time of its completion under the existing investment screening regime under the Enterprise Act 2002, it declined to do so despite sustained political pressure. Three successive decision-makers, including two Secretaries of State and the Prime Minister’s National Security Adviser, had found no grounds to act. The House of Commons Foreign Affairs Committee in March 2022 expressed its concern about ‘the Government’s apparent lack of appetite’ to use its screening powers to intervene.
Second, the timing of the call-in suggests that considerations unrelated to a genuine UK security assessment may have played a role. In April 2022, Members of the US Congress wrote to President Biden, urging the administration to engage with its UK counterparts ‘to secure disapproval of [the] acquisition’ and, if such diplomatic efforts were to prove unsuccessful, reconsider the UK’s status on the CFIUS whitelist as well as impose export controls on NWF. The transaction was called in six weeks later. While the timing may be circumstantial, it raises the question of whether the decision was driven by external political pressure.
Third, the security grounds articulated by the Secretary of State in the Order provided scarcely any context as to the specific risks to national security which the transaction gave rise to. According to the Secretary of State, the risks related to ‘technology and know-how that could result from a potential reintroduction of compound semiconductor activities at the Newport site, and the potential for those activities to undermine UK capabilities’ and that ‘the location of the site could facilitate access to technological expertise and know-how in the South Wales [Industrial] Cluster’, potentially making it harder for businesses in that Cluster to participate in future projects relevant to national security.
Fourth and finally, Nexperia claimed to have offered extensive remedies to address whatever security concerns the government may have had in relation to the acquisition – offers which, according to Nexperia, were completely ignored. However, section 26(3) of the NSIA requires a final order to be necessary and proportionate for the purpose of preventing, remedying or mitigating the identified risk to national security. If true, the rejection of mitigation offers and lack of engagement from the government’s side would further reinforce the view that the outcome of the screening process was not necessarily determined by the actual circumstances on the ground.
Taken together, the circumstances surrounding the Order against Nexperia raise questions as to whether the Secretary of State’s decision was genuinely driven by security considerations. As neither the NSIA nor any other UK legislation or guidance defines the term ‘national security’ (Hogan Lovells, 2022; Kokkoris, 2024, pp. 154-156), the executive is largely unfettered in determining its scope and contents.
While Nexperia subsequently pursued judicial review of the Order (eeNews, 2023; UK Parliament, 2023) – the outcome of those proceedings is not known at this time, and Nexperia sold NWF to U.S. firm Vishay Intertechnology in 2024 – the prospects of success before domestic courts are uncertain. In a recent judgment concerning a comparable NSIA divestment order, the UK High Court confirmed that ‘[p]arliament has given the primary responsibility for assessing both the nature of the risks to national security and for [assessing] what is a proportionate response to such risks to the Secretary of State … [t]he court will necessarily accord great respect to that assessment’ (R (FTDI Holding Ltd) v Chancellor of the Duchy of Lancaster [2025] EWHC 1922 (Admin), para. 112). Domestic judicial review – at least in the UK – thus offers investors little more than a procedural safety net.
Yet where the investor holds rights under international investment law, the fact that domestic courts are likely to defer does not mean that the exercise of screening discretion is legally unconstrained. The relevant question then becomes whether a screening decision, despite its national security framing, was taken in a manner consistent with the host State’s obligations under the applicable IIA.
III. The Outer Limits of Screening Discretion under International Investment Law
Wide, undefined screening discretion carries an inherent legal risk: it may permit decisions to be taken without genuine security justification, applied inconsistently, or driven by considerations entirely extraneous to the purpose for which the power was conferred. International investment law does not leave that risk unaddressed. While a range of treaty standards may come into play in the context of investment screening – including non-discrimination, protection against expropriation, and Fair and Equitable Treatment (“FET”) (see Voon and Merriman, 2022, p. 93; Abel, 2026, pp. 49-52) – it is the prohibition on arbitrary conduct that speaks most directly to the pathology of unconstrained discretion.
That prohibition is not confined to any single treaty standard. In some IIAs it operates as a standalone provision together with the prohibition of discrimination (see, e.g., Article II.2(b) of the United States-Argentina BIT). In others it appears as an element of the FET standard – codified, for instance, as an explicit prohibition of ‘manifest arbitrariness’ in Article 8.10(2)(c) of CETA. It is also broadly accepted as forming part of the minimum standard of treatment under customary international law (Stone, 2012, p. 92; Dolzer and others, p. 240). Across all of these formulations, the substantive content of the prohibition is broadly consistent, and it is against that standard that the exercise of screening discretion can be addressed.
IV. From Discretion to Arbitrariness
In order to establish the meaning of arbitrariness, several arbitral tribunals have utilised dictionary definitions of the term. For example, in Lauder v. Czech Republic (Final Award, 2002, para. 221), the Tribunal referenced Black’s Law Dictionary, according to which arbitrary means ‘depending on individual discretion; (…) founded on prejudice or preference rather than on reason or fact’. The Tribunal in Azurix v. Argentina (Award, 2006, para. 392), also relying on a dictionary definition, held that ‘[i]n its ordinary meaning, ‘arbitrary’ means ‘derived from mere opinion’, ‘capricious’, ‘unrestrained’, ‘despotic.’ […]’. The types of State measures found to meet this threshold are varied – including the blocking of payments, the transfer of contractual rights from a foreign to a domestic entity, and the permanent suspension of a foreign investor’s operations (Stone, 2012, p. 97).
When determining if a given State measure was arbitrary, several tribunals have examined whether it was the result of a rational decision-making process. In Saluka v. Czech Republic (Partial Award, 2006, para. 460), the Tribunal stated that ‘[t]he standard of ‘reasonableness’ therefore requires […] a showing that the State’s conduct bears a reasonable relationship to some rational policy […]’ (see also e.g., Plama v. Bulgaria, Award, 2008, para. 184; Lauder v. Czech Republic, Final Award, 2002, para. 232). According to the Tribunal in Micula v. Romania I (Award, 2013, para. 525), ‘the determination of whether the state’s conduct is reasonable requires the analysis of two elements: the existence of a rational policy; and the reasonableness of the act of the state in relation to the policy’. For a government action to be reasonable there must, in turn, exist an ‘appropriate correlation between the state’s public policy objective and the measure adopted to achieve it’.
Read together, these authorities suggest that arbitrariness is most likely to arise where a state acts without a genuine policy basis, without an adequate factual foundation, or without a reasonable relationship between the objective invoked and the measure adopted to achieve that objective. Translating this logic to investment screening, the consequences are that decisions taken without sufficient grounds to suggest that there is an actual risk to national security in the individual case, as well as decisions which are formally taken on security grounds but which are in reality intended to achieve other objectives, e.g., related to industrial policy, or to simply appease allies, run a substantial risk of violating the prohibition on arbitrary conduct.
To date, the only arbitral tribunal to have assessed whether the conduct of an investment screening authority was arbitrary is the one in Global Telecom v. Canada (Award, 2020). The claimant argued that Canada’s national security review violated the FET standard in Article 2(2)(a) of the Canada–Egypt BIT, as it was unreasonable, arbitrary, non-transparent and denied due process (para. 573). The Tribunal rejected that claim, finding that the evidence before it satisfactorily demonstrated genuine national security concerns assessed by the competent authorities (para. 616). Given that the award is heavily redacted, the precise nature or content of that evidence cannot be ascertained (see also Voon and Merriman, 2022, p. 100). What can be said, however, is that the Tribunal applied the arbitrariness standard directly to a screening decision, confirming that such decisions are in principle reviewable.
Turning to the NWF decision, any challenge by Nexperia under international investment law would fall to be assessed under the UK–China BIT. Notably, Article 2(2) of that treaty provides both for the FET standard as well as a standalone prohibition on ‘unreasonable or discriminatory measures’ – either of which could, in principle, serve as a basis for the arbitrariness argument developed above, mirroring the approach taken by the claimant in Global Telecom. Such a claim might, however, run into jurisdictional challenges. Article 7(1) of the UK-China BIT limits ISDS to disputes ‘concerning an amount of compensation’, a formulation whose precise scope remains contested in the arbitral case law on Chinese BITs (see Castro de Figueiredo, 2025).
Conclusion
Foreign investment screening regimes are, by design, built on discretion, but as this post has sought to demonstrate, that discretion is not without limit under international investment law. The prohibition on arbitrary conduct provides a principled basis for scrutiny where a screening decision lacks genuine security justification, is applied inconsistently, or is responsive to considerations unrelated to the purpose for which the power was conferred. The NWF case, assuming that the reported information holds true, would potentially display several of these characteristics and is therefore instructive as to the type of situation that may trigger investment treaty claims. As the proliferation of screening regimes continues, the pressure to test them legally, including through investment arbitration, will only grow.
