By Shameen Prashantham

Earlier this year, the social media and short-form video app TikTok, owned by Chinese internet technology company Bytedance, closed a deal to continue operating in the US, following years of negotiation between Beijing and Washington. The deal establishes a US-based joint venture structure, ensuring its continued access for millions of American users while addressing the political and regulatory concerns that have long caused tension.

Drawing on his research into asymmetric partnerships between businesses, CEIBS Professor of International Business and Strategy Shameen Prashantham looks beyond the headlines to examine what this joint venture really represents for multinationals in a complex geopolitical era, and how global companies operate in an era of “in-between” globalisation.

TikTok, the social media and short-form video app which is vastly popular around the world but owned by Chinese internet technology company Bytedance, has long been an issue of political debate in the US. Following prolonged uncertainty and scrutiny culminating in a 2024 law requiring divestiture, agreement was finally reached in early 2026 to create a new entity – TikTok USDS Joint Venture LLC. The venture will be 80.1% owned by American and global investors, with China’s ByteDance retaining 19.9%. Key managing investors include Oracle, Silver Lake and MGX, with US user data and the recommendation algorithm secured in Oracle’s US cloud.

There is an old saying that if you have a hammer, everything looks like a nail. Having spent two decades studying asymmetric partnering between large corporations and small startups, I am mindful of that risk. Yet the newly announced US joint venture around TikTok appears to be not merely a divestiture story, but a high-stakes experiment in managing asymmetry under geopolitical constraint. While on the surface it looks like a political compromise, it can be analysed as a striking case of asymmetric partnering. Not between a corporation and a startup, in this case, but among dissimilar actors spanning regulatory regimes and technological architectures.

In my book Gorillas Can Dance, I argue that the very asymmetries that make partnerships attractive often make them difficult to manage. Goal differences, structural misalignment and attention imbalances must be addressed deliberately if collaboration is to create value rather than friction. The TikTok case plays out these dynamics on a geopolitical stage.

It also reflects something deeper: a liminal moment in globalisation. When COVID struck, we experienced liminality, a “between-and-betwixt” reality between separate stable situations. In the years since, the geopolitical environment has remained volatile. This liminality is reflected in the TikTok joint venture, which sits between integration and decoupling, between Chinese technological capability and American political sovereignty. It is neither full separation nor seamless continuity. It is an in-between structure, and its durability will depend on how well that in-betweenness is managed.

Drawing on my research, three facets of asymmetry – and the corresponding remedies – are reflected in TikTok’s US joint venture, all of which can help us better understand its characteristics, and vulnerabilities.

1) Goal asymmetry — and the need to clarify synergy

For ByteDance, TikTok is a global strategic asset powered by proprietary algorithms and data network effects. For US policymakers, it is a potential threat to national security and data sovereignty. For the new American investors, the objective is commercial viability and long-term value creation. These goals overlap, but they are not identical. One side seeks global integration; another seeks insulation. One prioritises technological continuity; another prioritises governance control.

The deal attempts to clarify the synergy – which is crucial for addressing goal asymmetry – at the outset. The US venture secures and retrains the algorithm on American data, operates the US app, and safeguards user information. ByteDance retains a minority stake and revenue streams elsewhere. In effect, the collaboration defines a bounded zone: operational continuity within a governance perimeter.

For practitioners, this may matter more than formal ownership percentages. In asymmetric partnerships, progress depends less on perfect alignment than on clearly articulated, bounded wins for each party, a pragmatic balancing of political and strategic interests that enables (possibly uneasy) coexistence without full convergence.

2) Structure asymmetry — and the importance of creating interfaces

To address the asymmetry in the organisational make-up of the actors involved, unlike a traditional acquisition, this arrangement creates a mediating structure: the joint venture itself. It functions as an institutional interface between ByteDance and the US regulatory and political system. Oracle’s cloud is not merely infrastructure; it is architecture-as-governance. Data localisation becomes both a technical safeguard and a symbolic reassurance.

In corporate–startup settings, well-designed interfaces absorb friction between dissimilar actors; poorly designed ones amplify it. Here, the joint venture becomes a geopolitical interface between Chinese corporate involvement and American regulatory oversight. This distinguishes the arrangement from conventional deals or routine foreign subsidiaries. The novelty lies in algorithmic and data sovereignty being engineered directly into the organisational architecture. The structure is deliberately liminal: not fully independent of ByteDance, yet not fully subordinate to it.

Liminal arrangements can be fragile because they resist neat categorisation. Their legitimacy must be continually reinforced. The central question is whether this interface can provide enough regulatory confidence while preserving enough technological integration to sustain competitive advantage.

3) Attention asymmetry — and the imperative for cultivating exemplars

In asymmetric collaborations, cultivating exemplars of operational continuity is critical to attracting attention from all relevant parties and building confidence. For the TikTok US joint venture, early product stability serves as such an exemplar. The absence of dramatic algorithmic disruption is particularly important: after the announcement, deletions briefly spiked and speculation about censorship circulated widely, yet engagement metrics quickly stabilised, with daily active users returning to roughly prior levels and average time spent on the platform remaining high.

This illustrates a paradox of resilience. Even though ownership structures changed, governance was modified, and political scrutiny was intense, user behaviour proved remarkably stable. Users, in other words, primarily care whether the product works. At the same time, attention asymmetry cuts both ways. Regulators, advertisers and political actors remain vigilant. In liminal arrangements, perception can shift rapidly. Managing interpretation becomes as important as managing infrastructure, not least because public trust is perhaps the scarcest resource in this case.

Of course, it is not a fait accompli that this TikTok joint venture will turn out to be an exemplar. Its long-term durability will depend not only on legal compliance or political endorsement, but on disciplined management of goals, structure and attention asymmetries in a liminal global economy. If it succeeds, it will demonstrate that carefully engineered partnerships can preserve both sovereignty and scale. If it falters, it will underscore how fragile “in-between” architectures can be.

Either way, the lesson is clear: in today’s world, global strategy increasingly means learning to dance in liminal space.

Lessons for multinational enterprises: Toward “geo-partnering”

It is tempting to frame this episode in binary terms. Who won? Who lost? Was this a strategic victory for China’s tech sector or a triumph of American regulatory sovereignty? In the current geopolitical climate, that framing is counterproductive. We may instead be entering an era of managed interdependence, in which firms and governments seek partial, bounded wins that all parties can live with.

The TikTok joint venture may represent an emerging template for geo-partnering: collaborative architectures for managed interdependence designed to preserve scale while accommodating political constraint. Rather than full decoupling, we see engineered separation – minority stakes, localised data silos, algorithmic retraining within national boundaries – embedded within continuing commercial integration. This suggests that China-US economic interdependence has not disappeared, but now rests on conditionality and intentionality.

The TikTok US joint venture is not merely a workaround to avoid a ban, but potentially a prototype for how global firms survive when integration and fragmentation coexist. Such arrangements are not neat; they are liminal by design and may become increasingly necessary in a fragmented global order.

For multinational firms, three important lessons emerge.

First, asymmetry must be managed, not eliminated. Divergent goals can coexist if synergy is clearly defined and bounded. Firms operating across politically sensitive markets must articulate precisely what value is preserved and what governance concessions are made.

Second, the structure of partnering arrangements matters. In a world of regulatory divergence, firms may need institutional interfaces – joint ventures, data localisation entities, regionally ring-fenced units – that mediate between global integration and local sovereignty. As in TikTok’s use of Oracle’s cloud, organisational architecture can function as architecture-as-governance and thus as strategic signaling.

Third, adeptness at operating under liminal conditions is vital. The era of a steadily liberalising global order appears to be over. Customers may prioritise functionality over the optics of ownership, but this cannot be assumed. Liminal arrangements demand ongoing cultivation of trust. Multinational enterprises must navigate overlapping sovereignties, shifting regulatory expectations and politicised technological infrastructures.

Shameen Prashantham is Professor of International Business & Strategy, and Associate Dean (Africa), at CEIBS. He is the author of Gorillas Can Dance: Lessons from Microsoft and Other Corporations on Partnering with Startups, and is best known for his work on partnering between large corporations and startups.