A film or television production is an unusual form of economic undertaking. It combines creative ambition, substantial financial commitment, contractual obligations, specialist labour, technology, physical assets and a sequence of interdependent activities whose successful completion may determine whether the entire investment can be realised. A production exists simultaneously as a work of authorship, a coordinated industrial operation, a collection of legal rights, a financial proposition and an asset intended for eventual exploitation. None of these dimensions can be understood entirely in isolation. The significance of a delay may depend upon the availability of a particular performer; the consequences of a contractual dispute may extend to distribution; a change in technology may affect both the production schedule and the rights associated with the finished work. The undertaking is therefore not merely a collection of insurable events. It is a system of relationships in which the consequences of one event may propagate through many others.
For insurers specialising in film and television production, this creates a distinctive intellectual challenge. The assessment of an individual risk is indispensable, but the risk itself exists within a larger arrangement of dependencies. The production budget, financing structure, schedule, cast, crew, locations, equipment, weather, contractual commitments, intellectual property, completion requirements and intended distribution are connected in ways that may change throughout the undertaking. A vulnerability that appears manageable when considered separately may become consequential when combined with other conditions. Conversely, a difficulty in one part of a production may be absorbed without material loss because the wider system possesses sufficient flexibility, alternative resources or contractual protection. The quality of underwriting depends upon understanding not only what may go wrong, but how the production is likely to respond when circumstances depart from the plan.
UNIVERSAL INTELLIGENCE® begins with this proposition: the proper object of understanding is the whole production, considered through the relationships that determine its resilience, continuity, completion and commercial viability. Artificial intelligence may extend the insurer's capacity to analyse these relationships, connect dispersed evidence and investigate scenarios that would otherwise be difficult to evaluate comprehensively. It cannot remove the uncertainties inherent in creative enterprise, nor can it establish the future with certainty. Its value lies in supporting a more coherent understanding of the undertaking before the insurer and the insured must make decisions whose consequences may be substantial.
THE INTERDEPENDENCE OF PRODUCTION RISK
Conventional risk assessment necessarily divides complex undertakings into identifiable categories. Property, equipment, liability, personnel, interruption, completion, intellectual property and other exposures may be examined through different policies, contractual arrangements, specialist disciplines and underwriting processes. Such distinctions are necessary: each form of cover has its own purpose, conditions, limits and exclusions. Yet the practical behaviour of a production does not follow these administrative divisions. A single event may have consequences across several categories, while a combination of individually modest problems may create a loss that no isolated assessment adequately anticipates.
Consider the significance of a delay. Its initial cause might appear limited to the unavailability of a location, the failure of equipment, adverse weather or the absence of a key individual. The eventual consequences, however, may depend upon the production's remaining schedule, the availability of replacement resources, contractual restrictions, the costs of remobilisation, the commitments of financiers and distributors and the extent to which work can be rearranged. A disruption becomes material through the system into which it occurs. Its consequences cannot always be inferred from the original event alone.
The same principle applies to legal and commercial uncertainty. A question concerning the authority to use a performance, a piece of music, an archive recording or a digital asset may initially appear to be a narrow clearance issue. If the relevant material is central to the production, however, uncertainty over its permitted use may affect editing, delivery, distribution, contractual compliance and the ability to exploit the finished work. The significance lies not solely in the disputed asset, but in the relationships between rights, obligations, production choices and commercial commitments.
Universal intelligence, in this context, is an approach to underwriting that seeks to understand how risks combine, propagate and alter the conditions of the entire undertaking. It requires the insurer to consider dependencies rather than merely inventories, the mechanisms through which losses develop rather than only their initial causes and the resilience of the production rather than simply its declared resources. Artificial intelligence may help assemble and compare evidence across these dimensions, but the interpretation of that evidence must remain grounded in the realities of production and the terms upon which risk is accepted.
FROM IDENTIFIABLE HAZARDS TO SYSTEMIC CONSEQUENCES
The most consequential production risks are not always the most conspicuous. Some arise from a clear physical hazard or identifiable event; others emerge from the interaction of assumptions that appear reasonable when examined separately. A tightly compressed schedule may depend upon several resources remaining available at precisely the right time. A budget may be viable only if a sequence of approvals and contractual commitments proceeds without material delay. An apparent contingency may be inadequate if several disruptions occur together or if the alternatives relied upon are not genuinely available. Such vulnerabilities may remain obscured until the production encounters circumstances that expose their interdependence.
A sophisticated intelligence capability should therefore examine the conditions upon which a production depends. Which assumptions are essential to completion? Which resources have no practical substitute? Where is there schedule flexibility and where does one delay create further delays elsewhere? Which contractual obligations become critical if the production changes direction? How far does financial resilience depend upon a particular delivery date, financing arrangement or distribution commitment? These questions are not a substitute for conventional underwriting analysis. They reveal the wider system within which that analysis must be interpreted.
Artificial intelligence can support this work by connecting information from budgets, schedules, contracts, production reports, asset records and other appropriately available material. It may help identify inconsistencies, expose dependencies, compare alternative scenarios and draw attention to information that would otherwise remain dispersed across separate documents or departments. The strategic value is not simply the production of a longer risk report. It is the development of a clearer account of how a production's principal vulnerabilities relate to one another and which relationships deserve the greatest attention.
There remains an essential distinction between an identified dependency and a predictable loss. A system may reveal that several activities rely upon the same resource without establishing that the resource will fail. It may indicate that a delay could have significant downstream effects without determining whether those effects will materialise. Intelligence should make the structure of uncertainty more visible, not convert uncertain possibilities into apparently certain conclusions. Underwriting judgement remains necessary to determine which dependencies are material, what protections are proportionate and how the remaining uncertainty should be reflected in the terms of cover.
CREATIVE ENTERPRISE AND THE LIMITS OF PREDICTION
Film and television production cannot be reduced to an industrial process whose outcomes are wholly determined by its inputs. Creative quality, performance, audience response, artistic decisions and commercial reception contain uncertainties that are not resolved by more extensive analysis. A production may be technically completed and contractually delivered without achieving its anticipated commercial success; an unconventional creative decision may produce exceptional value despite appearing difficult to justify through historical comparisons. The distinction between operational completion and commercial achievement must therefore remain clear.
For insurers, this means that intelligence must be directed towards the risks that can reasonably be examined, rather than an imagined capacity to predict the entirety of a production's future. There may be considerable scope to improve understanding of operational dependencies, the credibility of budgets, contractual rights, contingency arrangements and the consequences of specified events. There will remain other questions for which evidence is incomplete, probabilities are unstable or the outcome depends upon creative and market forces that cannot be reliably modelled.
A disciplined framework distinguishes between what is known, what can be reasonably inferred, what can be tested through scenarios and what remains fundamentally uncertain. These categories should not be collapsed into a single confidence score. The apparent precision of a model can conceal the weakness of its assumptions, particularly where the historical record is limited or where a new production technique changes the nature of the exposure. A credible intelligence process should therefore disclose the basis of its conclusions, recognise the limitations of its evidence and retain the capacity to revise its assessment as circumstances develop.
This is not an argument for limiting the use of artificial intelligence. It is an argument for using it with greater discrimination. Its contribution may be greatest where it helps underwriting professionals understand complex dependencies, identify gaps in evidence and ask better questions. There are circumstances in which the most valuable result will be a finding that the available information does not justify a confident conclusion. Recognising the limits of knowledge is itself a form of intelligence.
THE CHANGING CHARACTER OF CREATIVE RISK
Artificial intelligence is changing the processes through which screen content can be conceived, produced, edited, modified and distributed. It can assist with research, translation, pre-visualisation, image generation, visual effects, voice technologies, synthetic environments and other activities within the production process. These capabilities may create legitimate creative opportunities and efficiencies, but they also alter the questions that producers, rights holders, distributors and insurers may need to examine.
The relevant issue is not whether artificial intelligence has been used in a production as a simple binary distinction between conventional and technological methods. Its significance depends upon the nature of the use, the material affected, the rights involved, the contractual arrangements, the production's intended exploitation and the evidence available to establish what occurred. An internal research application presents a different set of considerations from the generation of a principal performance, the reproduction of a recognisable voice or likeness, or the incorporation of material whose origin and permitted use cannot be established with confidence.
Rights, consent, authorship, provenance and contractual authority may become particularly important where machine-generated or machine-modified material forms part of a finished work. A production may depend upon underlying works, performances, recordings, images, datasets, software or third-party services. The capacity to generate new material does not itself establish that every input was lawfully obtained, that every resulting use is authorised or that every contractual requirement has been satisfied. Nor should the presence of artificial intelligence automatically be treated as evidence of infringement or uninsurability. The appropriate assessment depends upon the actual facts and the applicable policy wording, law and contractual arrangements.
The legal and commercial environment remains subject to development. In its March 2026 report on copyright and artificial intelligence, the UK Government addressed unresolved questions concerning the training of artificial intelligence systems, computer-generated works and digital replicas. Such developments illustrate why production intelligence must remain current and jurisdictionally aware, rather than relying upon assumptions that may become outdated. The underwriting task is not to substitute a general opinion about artificial intelligence for legal analysis, but to identify the relevant use, establish the available evidence, understand the rights and obligations involved and recognise where uncertainty requires further consideration.
PROVENANCE, CONSENT AND EVIDENCE
As production methods become more varied, the ability to establish the origin and authorised use of material may assume greater practical importance. A finished sequence may incorporate original performances, licensed material, generated imagery, modified recordings and contributions from several parties. A producer's ability to demonstrate that these elements were created or used within the relevant permissions may depend upon contractual records, licences, consent documentation, production logs and reliable information about the tools and processes involved.
The underlying principle is not that every production must generate an exhaustive record of every technical operation. It is that the evidence retained should be proportionate to the materiality of the use and the rights or obligations it may affect. Where a digital replica of a performer is involved, for example, the scope of consent, the specified use, the contractual basis and the security of the associated material may be relevant. Where generated content incorporates or resembles pre-existing work, the origin and permissible use of that material may require examination. Where a third-party technology provider is involved, the production may need to understand the terms upon which the system is used and the responsibilities allocated between the parties.
Artificial intelligence can assist with organising documents, comparing contractual requirements with available records, identifying missing information and maintaining a more accessible account of material decisions. It cannot, by itself, make an unauthorised use lawful, cure an absent consent or establish the truth of an unreliable record. Documentation is evidence to be evaluated, not a guarantee that the underlying position is sound.
For insurers, the implications extend to the evaluation of risk and the handling of claims. Clear records can help establish what was created, who authorised its use, what representations were made and which contractual obligations applied. Conversely, incomplete or contradictory records may make it more difficult to understand the facts, allocate responsibility or assess the extent of a loss. An intelligence-led approach seeks to identify such weaknesses before they become material, without assuming that documentation alone removes the underlying exposure.
THE INSURABILITY OF THE WHOLE
Insurance does not eliminate uncertainty. It establishes a defined contractual response to specified risks, subject to the terms, conditions, limits and exclusions of the policy. The quality of underwriting depends upon a disciplined assessment of the undertaking, the exposures presented, the available evidence and the relationship between the risk accepted and the terms on which it is accepted. Greater intelligence can improve that assessment, but it cannot replace the need to define the insured exposure with precision.
A production may contain numerous risks, but not every risk is necessarily insurable under every policy and not every uncertainty can be transferred to an insurer. Some exposures may be addressed through prevention, contractual allocation, contingency planning or operational controls; others may remain outside the agreed cover. The purpose of intelligence is to clarify these distinctions, identify where the assessment depends upon assumptions and improve the understanding upon which underwriting decisions are made.
This requires a clear relationship between analysis and policy language. A sophisticated risk assessment has limited practical value if it is not translated into an accurate understanding of what is being insured, under what circumstances cover may respond and which material limitations apply. Equally, a policy cannot be understood in isolation from the production it covers. The insured undertaking may change; assumptions may become invalid; new technologies may be introduced; and material information may emerge between inception and completion. The significance of these developments depends upon the policy terms and the circumstances of the case, but an intelligent underwriting process should be capable of identifying when reconsideration is warranted.
Universal intelligence therefore does not imply universal coverage, universal prediction or the elimination of exclusions. It represents a broader capacity to understand the production, its dependencies and the terms through which risk is accepted. Its purpose is to enable insurers to make better-informed decisions about the risks they undertake, the information they require and the conditions necessary for a defensible assessment.
THE UNDERWRITER AND THE INTELLIGENCE SYSTEM
Underwriting remains an exercise in judgement. It requires more than the calculation of probabilities or the extraction of patterns from historical information. The underwriter must interpret the character of the production, evaluate the credibility of its plans, understand its contractual and operational environment, assess the quality of the evidence and determine how the identified exposures should influence the decision. These tasks require experience, contextual understanding and a willingness to question conclusions that appear plausible but are insufficiently supported.
Artificial intelligence can strengthen this work by widening the field of inquiry. It can help the underwriter compare information from different sources, identify inconsistencies, examine alternative scenarios and bring relevant knowledge to bear more rapidly. It may also help preserve the reasoning behind earlier decisions, allowing the institution to understand how comparable risks were assessed and what experience subsequently revealed. The objective is not to turn underwriting into a mechanical calculation, but to provide professionals with a more complete intellectual environment in which to exercise their judgement.
The relationship must nevertheless be governed. A model may be influenced by incomplete historical data, may misinterpret an unfamiliar production arrangement or may give undue weight to variables that are easy to quantify. Its recommendations may appear precise despite uncertainty in the underlying evidence. If underwriters defer to such outputs without understanding their limitations, automation can reinforce error rather than improve judgement. Conversely, a disciplined process that compares model outputs with source evidence, preserves independent challenge and allows experienced professionals to question assumptions may produce material benefits.
The most capable underwriting institution will not necessarily be the one that automates the largest proportion of its decisions. It will be the one that determines where machine intelligence adds genuine value, where human expertise is indispensable and how responsibility should be allocated between them. The principle is proportionate autonomy: the scope of automated action should reflect the consequences of error, the reliability of the evidence and the institution's capacity to monitor, explain and correct its decisions.
PORTFOLIO INTELLIGENCE AND ACCUMULATION
The individual production is only one level at which an insurer must understand risk. Across a portfolio, apparently separate productions may share dependencies upon the same locations, specialist suppliers, technical infrastructure, distribution arrangements, professional services or broader economic conditions. Several insureds may be affected by a common event, or multiple claims may arise from a shared technological or legal development. Such accumulation may not be apparent when each production is considered independently.
Artificial intelligence can assist in examining these commonalities by connecting information across insureds, categories of exposure, policy periods and relevant external conditions. It may help identify concentrations, explore alternative scenarios and indicate where the insurer's portfolio depends upon assumptions that should be examined more closely. The objective is not to assume that all productions sharing a characteristic will suffer the same loss, but to improve the institution's understanding of potential common causes and the circumstances under which independent exposures may become correlated.
This is particularly relevant when production methods or technologies change rapidly. The adoption of a common third-party system, the use of similar contractual approaches or a change in the interpretation of a relevant legal obligation may create exposures that extend across a portfolio. The actual consequences will depend upon the facts, terms and jurisdictions involved, but the possibility of shared exposure makes portfolio-level intelligence an important complement to individual underwriting.
A more comprehensive view can also support institutional learning. Claims, near misses, risk improvements, underwriting decisions and changes in production practice provide evidence about the assumptions on which the portfolio has been built. Properly analysed, this experience can help the insurer distinguish recurring patterns from isolated events, understand where its methods require refinement and improve the consistency of its decisions. The purpose is to develop a more coherent understanding of the portfolio, not to allow historical patterns to dictate responses to unfamiliar circumstances.
INSTITUTIONAL MEMORY AND CONTINUOUS LEARNING
An insurer's intelligence is partly determined by what it can remember and how effectively it can use that memory. Over time, an institution accumulates knowledge through underwriting files, claims, policy interpretation, loss analysis, broker relationships, production experience and specialist judgement. Yet the existence of records does not guarantee that the knowledge within them remains accessible or that lessons from previous experience inform current decisions.
Artificial intelligence may assist in making institutional knowledge more usable. It can help locate relevant precedents, compare the circumstances of different losses, identify recurring weaknesses and connect historical decisions with subsequent outcomes. It can also help reveal where the organisation repeatedly encounters the same uncertainty, relies upon similar assumptions or lacks adequate evidence to assess a particular exposure. The value is not in retaining every historical document without distinction, but in preserving and retrieving the knowledge that can improve future judgement.
Learning must, however, remain intellectually disciplined. A loss does not necessarily prove that the original underwriting decision was unreasonable, just as the absence of a loss does not establish that an exposure was properly understood. Outcomes must be interpreted in relation to the information available at the time, the terms of cover, the assumptions made and the uncertainties that could reasonably have been recognised. Otherwise, the institution risks allowing hindsight to distort its evaluation of professional judgement.
A learning insurer uses experience to improve the questions it asks, the evidence it requires and the way it understands risk. It is capable of revising its assumptions without abandoning sound principles whenever an outcome differs from expectation. Artificial intelligence can accelerate the examination of accumulated experience, but the creation of institutional knowledge requires more than technical retrieval. It requires the willingness to challenge established practice and the discipline to distinguish what the evidence genuinely demonstrates from what it merely suggests.
UNIVERSAL INTELLIGENCE® AT GENERAL INTELLIGENCE PLC
UNIVERSAL INTELLIGENCE® is an artificial intelligence consultancy framework provided by X from GENERAL INTELLIGENCE PLC to insurance providers specialising in film and television production. It considers the strategic, technical and institutional implications of applying artificial intelligence to an industry in which creative work, financial commitments, legal rights, production operations and commercial delivery are deeply interdependent. Its defining concern is the development of a more comprehensive understanding of the production as a whole, rather than the isolated optimisation of individual processes.
The framework examines how intelligence can support the assessment of production dependencies, the interpretation of risk information, the evaluation of emerging technological exposures, the understanding of contractual and rights-related uncertainty and the development of more effective institutional learning. It also considers the relationship between underwriting judgement and machine capability, including the quality of the evidence required, the limitations of predictive systems and the governance necessary when automated analysis informs consequential decisions.
Its purpose is not to suggest that every production can be made predictable, that every uncertainty can be quantified or that the deployment of artificial intelligence will necessarily improve underwriting outcomes. Rather, it provides a structured way to consider where greater intelligence may improve the assessment and management of risk, where it may reveal weaknesses in existing assumptions and where human judgement must remain decisive. The relevant application will depend upon the insurer, the production, the nature of the exposure and the objectives for which the technology is being considered.
The word universal should be understood in this framework as an aspiration towards a whole-system perspective, not a claim to omniscience. The aim is to connect the elements of a production sufficiently well to understand how they interact, how risk may propagate and where the limits of available knowledge remain. The framework's commercial identity is distinct from the wider intellectual history of the terms universal and intelligence. Its substance lies in their particular application to production insurance and the institutional questions that application raises.
THE STANDARD OF UNIVERSAL INTELLIGENCE®
The test of UNIVERSAL INTELLIGENCE® is whether an insurer can develop a more complete and defensible understanding of the productions it insures. Can it identify the dependencies on which successful completion relies? Can it recognise how an apparently local event may affect the wider undertaking? Can it distinguish documented fact from assumption and a plausible scenario from a reliably established conclusion? Can it evaluate the implications of artificial intelligence for rights, consent, provenance and contractual authority without treating every use as either inherently acceptable or inherently unacceptable? Can it translate analytical findings into underwriting decisions and policy terms that reflect the actual exposure? Can it learn from claims and experience while remaining alert to risks that have no reliable historical precedent?
The test also extends to institutional conduct. Can the insurer use increasingly sophisticated tools without surrendering the independent judgement required to assess their outputs? Can it recognise when the evidence is insufficient? Can it understand the common exposures that may affect an entire portfolio? Can it maintain records, governance and accountability proportionate to the decisions being made? And can it improve its understanding without promising a degree of certainty that the nature of creative enterprise does not permit?
These questions define an ambition that is both demanding and practical. Intelligence should improve the quality of the underwriting process, strengthen the interpretation of risk and make the institution more capable of learning. It should not be measured by the volume of analysis produced, the complexity of the models used or the confidence with which an uncertain conclusion is expressed. Its value lies in the understanding it creates, the judgement it supports and the quality of the decisions that follow.
THE INSURABILITY OF CREATIVITY
Film and television will continue to combine human imagination with industrial coordination, financial exposure and technological change. The methods of production will evolve, the rights and obligations associated with creative work will develop and new forms of uncertainty will emerge. Yet the fundamental task of production insurance will remain recognisable: to assess defined risks, establish the terms upon which they can be accepted and provide a disciplined contractual response when covered events occur.
Artificial intelligence may improve the capacity to understand the circumstances in which these risks arise. Its contribution will be greatest where it enables the insurer to see relationships that would otherwise remain obscure, to identify material assumptions before they fail and to learn more effectively from the experience of successive productions. But insurance will continue to require judgement about what can be known, what can be managed, what can be transferred and what uncertainty must remain outside the protection offered.
UNIVERSAL INTELLIGENCE® provides a framework for approaching that task. It treats the production not as a collection of isolated risks but as a connected undertaking; not as a system that can be made entirely predictable, but as one whose dependencies can be understood more rigorously; and not as a technical problem alone, but as a combination of creative, financial, legal, operational and institutional questions. Its purpose is to improve the intelligence with which production risk is assessed, while preserving the professional judgement and contractual discipline upon which insurance depends.
The enduring objective is not to eliminate uncertainty from creativity. It is to understand the structure of that uncertainty more fully, to assess its consequences more intelligently and to make decisions with greater discipline about the risks that can be insured. In an industry where the whole may be materially different from the sum of its parts, the capacity to understand the whole is a strategic capability in its own right.