The ‘Integration Gap’ Problem: Funding Tools, Ignoring Systems
Amid escalating global political tensions and mounting hostility toward scientific communities, one reality stands out: the world is undergoing a massive sustainability transition.
In the UK, promising startups are developing novel, biodegradable materials designed to replace single-use plastics in mainstream retail supply chains, with some major brands already integrating the technology into their packaging systems. The renewable energy transition has also accelerated dramatically: in 2024, renewable sources supplied a record 50.4% of UK electricity generation.
Meanwhile, artificial intelligence (AI) has shown potential to improve company performance in sustainability, with tools being purposely built to achieve this goal. Although AI’s rapid scaling has drawn scrutiny: the United Nations warns of the growing energy and resource demands associated with large-scale AI deployment.
Despite the radical promise across these sectors, systems have been sluggish to adapt, and people working to curb climate change are growing impatient.
The problem reflects something more longstanding and structural: a recurring integration gap between technological breakthroughs and the infrastructures, institutions, and organisational systems required to accomodate them at scale.
Historical and theoretical lenses
The integration gap is not a new concept, and can be identified throughout major innovation shifts in recent history.
By the late 19th century, electricity generation became viable, but widespread electrification took decades. A U.S.-based study notes that moving from 10% to 90% household electrification took around 40 years (from Edison’s first grid in 1882 to near-total adoption by the 1920s) and was only achievable through aggressive government planning and investment.
Similarly, the Internet was commercially deployed by the mid-1990s, but meaningful productivity gains in the economy did not appear until the 2000s – this was coined as the “productivity paradox” by MIT professor Erik Brynjolfsson.
The historical cases of full-scale electrification and internet deployment show how new technologies can take decades to permeate the mainstream, due to misalignment between investment initiatives and regulatory & systemic change.
Sociologist Frank Geels’ Multi-Level Perspective (MLP) captures this notion of transition theory effectively: radical innovations flourish in protected niches but typically stall when confronting the inertia of established regimes (incumbent industries, regulations, standards) that are ‘entrenched and stabilised by lock-in mechanisms’.
Finally, the landscape layer accounts for all other major external events that impact society, such as armed conflict, pandemics, and economic crises. For the remainder of this article, we will use the MLP lens to interpret different emerging innovations.
The Circular Economy Problem
The circular economy (CE) – a concept centred on closing material loops and eliminating waste – has been extensively explored across scientific, policy, and business communities, generating a wide range of niche innovations, from recyclable biomaterials and reuse schemes to take-back services and reverse logistics platforms.
Yet widespread adoption remains limited due to deeply entrenched regime-level barriers.
Firms face significant organisational inertia when attempting to restructure linear production systems, while upstream suppliers often resist transition because their business models depend on decades-old one-way resource flows.
At the same time, fragmented and inconsistent regulatory frameworks – both nationally and across international markets – in addition to broader socio-political trends further complicate the implementation of circular systems.
More fundamentally, modern supply chains were designed for the existing regime of linear throughput, not material recirculation. Circular systems rely on reverse logistics networks capable of processing unpredictable volumes, timing, and quality of returned goods – conditions that traditional supply-chain models struggle to accomodate efficiently.
As a result, even highly promising innovations frequently remain confined to pilot projects and niche markets. Recent research shows that CE technologies continue to cluster within isolated demonstration projects and dense metropolitan regions, while less connected areas often lack the infrastructure, coordination mechanisms, and policy support necessary for wider diffusion.
The outcome reinforces a familiar pattern: the innovation itself may function, but the surrounding regime remains structurally incompatible for large-scale integration.
Renewable Energy: Pattern Confirmed
Renewable energy has already long-passed the niche-level phase. In terms of cost efficiency, renewables have already established a decisive advantage over fossil fuels: in 2024, roughly 90% of new utility-scale renewable projects globally were cheaper than the least-cost fossil-fuel alternative, with solar energy estimated to be 41% cheaper and wind 53% cheaper.
The technological breakthrough, in other words, has largely already occurred. The bottleneck still lies in the regime built to accommodate it. Modern electricity grids were designed around centralised fossil-fuel generation, and as a result, grid integration for renewables has become the defining constraint.
In the United States, around 80% of proposed renewable projects withdraw before connecting to the grid because connection delays and upgrade costs become economically unjustifiable; the average time spent in interconnection queues rose by ~70% over the last decade.
In the UK, the picture is similar: companies today may wait up to 15 years for a grid connection, and the pipeline of projects has grown tenfold in the last five years.
This shows the constraint does not lie in generation (we have more renewable capacity than ever) but in transmission infrastructure, permitting, and grid-management systems designed for an entrenched energy landscape. Enormous investments are pouring into clean energy – but there is a significant integration gap between the niche and the regime. In other words, like CE, this is not a funding shortage but an infrastructure absorption problem.
AI for Sustainability
Now consider AI in the enterprise. Although adoption is accelerating rapidly, organisational readiness to absorb it remains limited. McKinsey reports that between 78–88% of companies now use AI in at least one business function, yet only around 1% describe themselves as truly ‘AI-mature’ – meaning they have the infrastructure and confidence to integrate AI across core operations.
The result is a growing gap between deployment and meaningful transformation.
A recent MIT study similarly finds that 95% of generative AI pilot projects fail to deliver measurable ROI, largely due to insufficient data readiness, governance frameworks, and workforce capability rather than model performance itself.
In the UK, only one in six firms have embedded AI into regular operations, with most citing limitations in data infrastructure, skills, and organisational readiness.
AI resembles the initial stages of CE and renewable energy transitions, but importantly it reflects a highly unique case where niche-level innovation has scaled at an exponentially faster pace than the surrounding regime.
Crucially, AI’s pace has also drawn intense scrutiny regarding its immediate environmental impact in a time when our window to address climate change is getting narrower: vast amounts of water and electricity are required to operate current data centres, and global communities (especially those who suffer most from climate disasters) are becoming increasingly concerned.
The critical risk, therefore, is that organisations deploy AI at surface level – optimising reporting and visibility – without any underlying process transformation, and embracing a form of greenwashing rather than systemic change.
Breaking the Cycle: Policy and Systems Solutions
Each of the three innovation pathways mentioned in this article finds itself in a different regulatory, social, and technological context. But what they all have in common is a critical integration gap, and as exemplified by the MLP framework, it is a normal process of human-induced change.
To unlock the full transformative potential of each sector, actors across the niche, regime and landscape levels must collaborate and treat system integration as a coequal goal.
This means funding research along with integration infrastructure – e.g., electrical grid upgrades, reverse logistics and recycling networks for CE, and data governance and interoperability standards for AI.
Governments can drive this through better-targeted procurement and regulation initiatives (e.g., requiring open data standards, circular design norms, or grid connection reforms). The UK has begun to recognize this: new government guidelines explicitly warn that “the effectiveness, safety and legitimacy of AI adoption remain fundamentally constrained by the quality, structure and governance of underlying data”, and they call for making public datasets “AI-ready” with clear data governance.
The UK also aims to improve energy connection queues by prioritizing actively deployable projects and weed out outdated ones to unclog the system, in addition to creating coalitions that bring critical CE stakeholders under a single cooperative umbrella.
However, it is AI – the most heavily capitalised of these innovations in recent times – that serves as the ultimate test of this transition. While billions are poured into model development, the integration gap remains its greatest bottleneck.
If we treat AI as merely a greenwashing mechanisms for inefficient, linear processes, we risk a new level of environmental harm. To avoid the decades of lag seen in innovation cases such as electrification, we must pivot from funding the intelligence in isolation to building the systemic readiness that allows it to function.
The entrenched regime won’t transform itself; it necessitates coordinated pressure from policy, markets, and society. It is essential to align innovation funding with radical regulatory reform, ensuring that our most powerful innovative tools are both deployed and integrated.
