Film and television production represents an unusually sophisticated combination of creative ambition, financial commitment and operational interdependence. A production must bring together the contributions of numerous individuals and organisations, coordinate resources within defined periods, satisfy contractual obligations, establish the necessary rights and permissions and ultimately deliver a work capable of being exploited commercially. It must do so in circumstances where important events cannot be guaranteed, where plans may change and where a difficulty in one part of the undertaking may have consequences far beyond its immediate point of origin. The production is therefore both an exercise in organised creation and an exposure to uncertainty.
The willingness to undertake such uncertainty is fundamental to the creative industries. Producers commit resources before the final work exists; financiers make decisions before its commercial reception can be known; distributors enter arrangements on the basis of anticipated delivery and exploitation; and insurers agree to bear defined risks in return for a premium. Each participant approaches uncertainty from a different perspective, with different responsibilities and different means of managing the consequences. Yet all depend upon some reasonable understanding of the undertaking, its vulnerabilities and the circumstances that may determine its success or failure.
Insurance occupies a distinctive position within this arrangement. It does not remove the possibility of loss, guarantee completion or establish the commercial success of a creative work. Its function is to provide a contractual response to specified events, subject to the agreed terms of cover. By allowing defined risks to be transferred, it can form an important part of the financial and operational arrangements upon which a production depends. Its effectiveness therefore relies upon more than the existence of a policy. It depends upon the quality of the assessment preceding the acceptance of risk, the clarity of the contractual terms and the institution's capacity to respond appropriately when the facts differ from expectations.
The future of production insurance will be shaped, in part, by the intelligence with which these conditions are understood. Artificial intelligence offers new possibilities for examining complex information, identifying dependencies, testing assumptions and learning from experience. Its strategic significance lies not in a promise to eliminate uncertainty, but in the possibility of making uncertainty more intelligible. The insurer that understands the structure of a production more completely may be better equipped to determine which risks can be accepted, which require further evidence, which may be mitigated and which cannot be accommodated on acceptable terms.
UNIVERSAL INTELLIGENCE® approaches this question as an institutional challenge. Its concern is the development of a more comprehensive understanding of creative enterprise and the conditions under which it can be insured. It brings together the connected character of production, the changing nature of technological and legal risk, the discipline of underwriting and the responsibilities of institutions that must make consequential decisions without possessing complete knowledge of the future.
The economic purpose of insurance
Insurance is sometimes understood principally through the payment of claims following a loss. That is an essential part of its function, but it does not exhaust its economic significance. Insurance also provides a mechanism through which defined uncertainties may be incorporated into the financial arrangements of an undertaking. Where cover is available on acceptable terms, a producer or financier may be able to proceed with a project while maintaining a clearer understanding of the consequences of specified adverse events. Contractual certainty can assist planning, support financing arrangements and establish a recognised allocation of responsibility between the parties.
In film and television production, this function can be especially consequential. Budgets may be committed well before a production reaches completion and obligations to cast, crew, facilities, suppliers, financiers and distribution partners can arise at different stages. The financial consequences of disruption depend upon the structure of these commitments, the flexibility of the schedule, the availability of alternatives and the terms governing the production. Insurance cannot make these dependencies disappear. Properly structured cover may, however, provide a defined financial response to certain losses arising from them.
The insurability of a production is consequently not determined by whether uncertainty exists. Uncertainty is inherent in the activity. The relevant question is whether the exposure can be understood sufficiently, defined contractually and accepted on terms that appropriately reflect its characteristics. Some risks may be supported by extensive historical experience and reliable information. Others may be understood only through the examination of specific circumstances, while new or unusually complex exposures may present difficulties that cannot be resolved merely by applying a familiar classification.
A sound insurance market requires the capacity to distinguish among these conditions. Excessive confidence may lead to risks being accepted without adequate understanding; excessive caution may prevent cover from being offered where a more complete assessment could support a defensible decision. Neither is an adequate objective. The purpose of underwriting intelligence is to improve the quality of the distinction: to understand what is being undertaken, establish the extent to which the exposure can be assessed and determine whether the proposed transfer of risk is commercially and contractually justified.
This is where artificial intelligence may contribute to the economics of insurability. By assisting with the analysis of production information, contractual dependencies, historical outcomes and alternative scenarios, it may enable underwriters to investigate exposures with greater breadth and consistency. The benefit, where demonstrated, would not be the disappearance of uncertainty, but an improvement in the evidence available when deciding how that uncertainty should be managed.
From the individual risk to the whole production
The production is the natural unit of creative work, but its risk cannot always be understood by examining each component independently. A budget, schedule, cast list, location plan, equipment inventory or collection of contracts provides information about a particular aspect of the undertaking. Its wider significance emerges through the relationships between these elements. A delayed scene may affect the availability of a performer; a change in location may require additional permissions and expenditure; a contractual restriction may constrain an alternative production method; and an alteration to the intended use of a performance may have implications for rights, delivery and exploitation.
These relationships are not simply administrative details. They can determine whether a production can absorb disruption, reorganise its activities or continue without a material deterioration in its financial position. A difficulty becomes consequential through the conditions into which it occurs. The same event may have modest effects in a production with adequate contingency and substantial flexibility, yet threaten completion where the schedule is compressed, resources are difficult to replace or contractual commitments leave little room for adjustment.
The task of intelligence is to reveal these connections. An insurer needs to understand not only which hazards exist, but how a material event might affect the undertaking as a whole. Which assumptions are essential to completion? Which resources have no realistic substitute? Where do several activities depend upon one individual, supplier, facility or technical capability? Which contractual obligations become critical if circumstances change? What financial consequences arise if a delivery milestone is missed? These questions help distinguish the presence of risk from its potential significance.
Artificial intelligence may assist by examining information that has traditionally been dispersed across documents, systems and professional disciplines. It can help compare budgets with schedules, connect contractual obligations with production dependencies, identify inconsistencies and explore scenarios in which several conditions change together. It may also reveal gaps in the evidence, including where assumptions about alternative resources or recovery arrangements have not been substantiated.
Nevertheless, identifying a dependency does not establish that it will fail and constructing a scenario does not establish that the scenario will occur. The purpose of whole-system analysis is not to transform every possibility into a forecast. It is to make the structure of the undertaking more visible, enabling the underwriter to distinguish material vulnerabilities from incidental complexity and to consider the evidence required for a defensible decision.
The boundaries of the insurable
Not every uncertainty should be treated as an insurance exposure and not every exposure can be transferred on acceptable terms. Some risks may be mitigated through better planning, contractual arrangements, alternative resources, technical controls or other preventive measures. Some may be capable of being insured only within defined limits, subject to exclusions or specific conditions. Others may remain insufficiently understood, commercially disproportionate or unsuitable for transfer. An intelligent assessment must recognise these differences rather than assume that every risk can be resolved by a more sophisticated model.
The distinction is particularly important in creative enterprise because the uncertainties associated with production are not all of the same kind. The failure of specified equipment, the unavailability of a key person, damage to physical assets, an interruption arising from a covered event and a dispute concerning intellectual property raise different questions of causation, evidence, contractual responsibility and financial consequence. They cannot be treated as interchangeable expressions of a general risk score.
There is also a fundamental difference between an event that prevents or disrupts production and an outcome that disappoints commercial expectations. A completed work may fail to achieve its anticipated audience or return; a commercially successful production may incur an unexpected cost; and a technically successful process may not produce the artistic result originally envisaged. The existence of a financial disappointment does not, in itself, establish an insured loss. The relevant response depends upon the policy wording, the circumstances, the applicable terms and the nature of the exposure accepted.
Intelligence must preserve these distinctions. Its purpose is not to expand the apparent scope of insurability without regard to the contractual foundations of the insurance relationship. It is to establish a clearer understanding of the risks presented, the evidence supporting their assessment and the limits within which a transfer of risk can responsibly be undertaken.
This discipline protects both sides of the insurance relationship. Producers require a clear understanding of what their cover is intended to provide; insurers need to understand the exposure they have accepted and the conditions upon which their obligations arise. A more intelligent assessment can help bring these expectations into closer alignment, reducing the likelihood that important assumptions remain implicit or that the practical needs of the production differ materially from the protection purchased.
Artificial intelligence and the changing object of insurance
Artificial intelligence is altering the methods through which screen content can be researched, created, modified, assembled and distributed. It may assist with pre-visualisation, translation, editing, image generation, visual effects, synthetic environments and the alteration or generation of voice and likeness. Its uses will vary significantly in nature and consequence. Some may improve existing workflows; others may change the allocation of creative work, the evidence required to establish rights, or the contractual relationships through which a production is financed and exploited.
For insurers, the important question is not whether a production has used artificial intelligence in some general sense. It is what has been used, for what purpose, under whose authority, using which materials and with what consequences for the insured undertaking. A tool used for internal research presents a different set of considerations from a system used to generate a principal performance, reproduce a recognisable likeness or incorporate material whose origin and permitted use are uncertain. The significance of the technology depends upon the circumstances of its application.
The legal and commercial environment is also evolving. In its report on copyright and artificial intelligence published on 18 March 2026, the UK Government examined the use of copyright works in the development of artificial intelligence systems and the policy options associated with the issue. The report made clear that further evidence and assessment remained important before decisions on reform. For production insurers, this is an illustration of the need to distinguish an emerging area of legal and commercial uncertainty from an established rule that can be applied without qualification. The relevant position may depend upon the jurisdiction, the particular work, the conduct involved, contractual arrangements and subsequent legal or policy developments.
The implications extend beyond copyright. A production may need to consider consent to the use of a performer's likeness or voice, contractual provisions governing digital alteration, the provenance of generated material, confidentiality restrictions, the terms of third-party technology services and the obligations owed to financiers or distributors. These are distinct considerations, but they may interact. An unresolved rights issue involving material central to a production may affect its ability to deliver the work, satisfy contractual representations or exploit the finished content as intended.
Artificial intelligence can help identify relevant documentation, compare requirements, reveal missing evidence and support the examination of contractual dependencies. It cannot make an unauthorised use lawful, establish consent where none was given or guarantee that an uncertain rights position will be resolved favourably. The appropriate objective is a more complete understanding of the actual use, the associated evidence and the potential consequences for the production.
The future of insurability will therefore depend partly upon the capacity to understand new forms of production risk without exaggerating or minimising them. Treating every use of artificial intelligence as inherently unacceptable would ignore differences in purpose and exposure. Assuming that every technological innovation is safe because it improves efficiency would be equally unsound. Intelligence requires a more exacting approach: examining the facts, identifying the relevant uncertainty and determining what evidence, controls or contractual arrangements are necessary.
Intelligence before commitment
The value of intelligence is often greatest before an institution becomes committed to a course of action. Once financing has been arranged, contracts have been entered into and production resources have been assembled, the cost of changing direction may be considerable. A vulnerability that could have been addressed at the planning stage may become difficult or expensive to resolve later. An unresolved contractual dependency may constrain available alternatives. An assumption about a technical process may prove more consequential when changing it would disrupt the schedule.
For insurers, this makes the quality and timing of pre-commitment understanding important. The objective is not to require an unattainable level of certainty before every policy can be written. It is to establish whether the information material to the particular exposure is sufficient for the proposed decision and to identify where additional evidence or a different allocation of risk may be required.
Artificial intelligence may support this task by examining a production's financial, operational and contractual information as a connected body of evidence. It could help identify inconsistencies between declared plans and underlying documentation, draw attention to dependencies that have not been adequately explained and explore the consequences of plausible disruptions. It may also help underwriters compare a production's arrangements with relevant experience from previous undertakings, while accounting for important differences rather than treating superficially similar cases as equivalent.
The decisive question remains whether the analysis changes the quality of the decision. An extensive report has limited value if it merely restates the information already supplied. A model adds little if its conclusions cannot be traced to reliable evidence or if its output does not affect the assessment of the actual risk. The objective is to direct attention towards matters that are material, identify uncertainty that could alter the underwriting decision and support a reasoned understanding of the conditions under which the risk can be accepted.
A more intelligent process may lead to different outcomes in different cases. It may support acceptance on clearly defined terms, justify additional information, identify measures that reduce exposure, or establish that the risk cannot be accommodated on acceptable terms. The value lies in the quality of the reasoning, not in a presumption that intelligence must always result in acceptance. A rigorous decision not to underwrite an exposure can represent a successful application of intelligence when the evidence does not support a defensible transfer of risk.
The discipline of underwriting judgement
Underwriting is not merely the calculation of a probability or the assignment of a price to an exposure. It requires the interpretation of evidence in relation to the particular undertaking, the assessment of uncertainty and the determination of the terms upon which the insurer is prepared to accept responsibility. Historical data, actuarial methods, technical analysis and predictive models may all contribute, but none removes the need to exercise judgement where the circumstances are incomplete, unfamiliar or changing.
Artificial intelligence can expand the underwriter's analytical capacity. It may assist in examining complex documentation, identifying relationships across sources, comparing different scenarios and making relevant institutional knowledge more accessible. It can help formulate questions that require further investigation and reveal where a production's declared assumptions are inconsistent with the available evidence. Such capabilities may be especially useful where the risk spans creative, operational, legal and financial disciplines that are not easily examined through one conventional method.
Yet the use of artificial intelligence introduces its own uncertainties. Models may rely upon historical experience that does not adequately represent a new form of production. The evidence may be incomplete, inconsistent or biased towards cases that differ materially from the one under consideration. A model may produce an apparently precise result without demonstrating that its assumptions are reliable. If an underwriter treats such an output as authoritative merely because it is computationally sophisticated, the system may reinforce false confidence rather than improve understanding.
A mature approach must therefore preserve a distinction between analysis, recommendation, authorisation and action. The system should be capable of contributing evidence and reasoning within a defined scope, while the institution remains clear about who is responsible for the decision and what authority has been delegated. The more consequential the decision, the more important it becomes to understand the reliability of the evidence, the limits of the model and the process through which the result may be challenged.
This is not an argument against greater automation. It is an argument for proportionate autonomy. Routine tasks supported by reliable evidence may be suitable for extensive automation; complex or consequential decisions may require additional examination and explicit human authority. The appropriate boundary depends upon the nature of the exposure, the reliability of the system and the consequences of error. Intelligence is demonstrated not by automating the maximum number of decisions, but by understanding which decisions can be automated responsibly and which require a different form of judgement.
From individual productions to portfolio intelligence
The insurer's responsibility extends beyond the production considered in isolation. A portfolio of insured productions may contain common dependencies that become significant when conditions change. Several productions may rely upon the same specialist supplier, facilities, geographical location, technology provider or category of professional expertise. A legal or contractual development may affect multiple insureds. An event that appears local when considered against one production may have wider consequences across a portfolio.
Portfolio intelligence seeks to recognise these commonalities without assuming that shared characteristics necessarily produce identical outcomes. Artificial intelligence may help examine exposure information across productions, identify potential concentrations, investigate common causes and explore scenarios that affect several insureds simultaneously. The resulting analysis can support a more informed understanding of accumulation and the possibility that seemingly distinct risks may respond to the same underlying event.
This becomes particularly relevant as technology changes production methods. A common digital service may introduce dependencies across many productions, while a change in the treatment of rights, the operation of a platform or the terms of a third-party service may affect the exposures of multiple insureds. The nature and extent of any resulting loss would depend upon the facts, policy wording and applicable law. The purpose of portfolio analysis is to identify where such shared exposure deserves attention, not to presume that a single development will lead to a common claim.
A broader intelligence capability may also improve the allocation of institutional attention. Underwriters and risk professionals need to understand which concentrations are material, where evidence is weak and which developments justify changes to assessment methods or underwriting appetite. Artificial intelligence can help bring these issues into view, but the institution must remain capable of distinguishing an important pattern from a statistical coincidence and a credible scenario from an unsupported worst case.
The objective is a portfolio whose structure is understood more clearly. This means recognising not only the risks accepted, but the relationships among them, the assumptions upon which they depend and the conditions under which those assumptions might fail together. The value of intelligence lies in improving the insurer's ability to see the aggregate exposure without losing sight of the particular circumstances of each insured production.
The insurer as a learning institution
An insurer accumulates experience through underwriting decisions, claims, disputed losses, risk improvements, changes in production practice and the subsequent performance of the risks it has accepted. Over time, this experience may reveal recurring dependencies, weaknesses in assumptions and new forms of exposure. Yet the existence of records does not guarantee that the knowledge within them will inform future decisions. Experience can remain dispersed across files, systems, teams and individuals, leaving the institution to rediscover lessons that it has already encountered.
Artificial intelligence offers the possibility of making this institutional memory more accessible. It can help retrieve relevant cases, compare the circumstances surrounding losses, identify patterns in claims experience and connect later outcomes with the reasoning used when the original risk was accepted. It may also support the examination of whether underwriting assumptions remain appropriate as technology, contractual arrangements and production methods change.
Such learning must be undertaken with discipline. A claim does not automatically demonstrate that the original underwriting decision was unsound, just as the absence of a claim does not establish that a risk was adequately understood. Outcomes must be assessed in relation to the information available at the time, the terms of cover, the assumptions made and the nature of the uncertainty. Otherwise, hindsight can distort the evaluation of decisions, encouraging institutions to confuse an unfortunate result with unreasonable judgement or favourable experience with sound risk selection.
A learning insurer should aim to improve both its knowledge and its methods. It should be able to recognise where previous assumptions have proved robust, where new information has exposed weaknesses and where a changing environment requires different questions. It should also recognise when the available evidence is insufficient to support a general conclusion. Institutional learning is not the automatic repetition of historical patterns; it is the disciplined reconsideration of experience in the light of changing circumstances.
The strategic significance is cumulative. An insurer that can preserve relevant reasoning, learn systematically from claims and connect experience to future decisions may improve its capacity to understand exposures over time. Artificial intelligence can accelerate the analysis, but institutional culture determines whether the results are examined critically and translated into better practice. The ultimate capability belongs not to the model alone, but to the organisation that learns through its use.
Governance, evidence and institutional confidence
As artificial intelligence becomes more deeply integrated into underwriting and claims processes, the insurer must understand the systems upon which it increasingly relies. This requires attention to data provenance, model limitations, security, confidentiality, third-party dependencies, permissions, auditability and the allocation of responsibility. The nature of the required controls will depend upon the function performed and the consequences of failure, but the underlying principle is constant: analytical capability must not outpace the institution's capacity to govern it.
The insurance sector is itself moving towards more structured governance of artificial intelligence. In April 2026, the Lloyd's Market Association reported that a survey of participating market firms found that 93 per cent had an artificial intelligence governance framework either in place or in development. The finding illustrates an institutional priority rather than proof that governance challenges have been resolved. The transition from experimental applications to consequential underwriting and claims use requires sustained attention to reliability, authority and control.
For production insurance, these questions are inseparable from the quality of evidence. A system may help summarise a contract, but the institution must retain the ability to inspect the underlying document where necessary. An automated assessment may identify a potential rights issue, but it does not replace an appropriate legal evaluation. A model may suggest a relationship between previous losses and present conditions, but the insurer must establish whether the comparison is relevant to the actual exposure. The credibility of intelligence depends upon the capacity to trace its conclusions to evidence and to understand how those conclusions were reached.
Governance should therefore be regarded as an enabling condition of intelligence, not merely an obstacle to innovation. Its purpose is to ensure that the institution knows what its systems are permitted to do, understands where their limitations lie and retains the ability to challenge their outputs. It should support appropriate innovation while preserving the accountability necessary to maintain confidence in underwriting decisions.
UNIVERSAL INTELLIGENCE® and the future of insurability
UNIVERSAL INTELLIGENCE® is an artificial intelligence consultancy framework provided by X from GENERAL INTELLIGENCE PLC to insurance providers specialising in film and television production. Its purpose is to consider how artificial intelligence may support a more comprehensive understanding of the productions being insured, the relationships through which risks develop and the institutional capabilities required to assess and manage those risks responsibly.
The framework treats production as a connected undertaking rather than a series of independent exposures. It considers the interactions between creative objectives, budgets, schedules, people, resources, contractual rights, technological processes, financing and commercial delivery. It also examines how emerging forms of production risk may affect the evidence required for underwriting, how intelligence can improve the assessment of dependencies and how accumulated experience can be converted into more effective institutional knowledge.
Its ambition is not to create an insurer that claims to predict every disruption or eliminate every uncertainty. Nor does it presume that greater analytical sophistication must lead to more generous cover, lower premiums or superior claims outcomes. Those results cannot be guaranteed. The framework instead provides a basis for examining whether greater intelligence can improve the quality of the underwriting process: whether the institution can understand more clearly what it is being asked to insure, identify material gaps in evidence, assess alternative scenarios and establish terms that reflect the actual exposure.
The framework also recognises the limits of technological capability. Artificial intelligence can help organise information, reveal relationships and support the formulation of judgements. It cannot, by itself, determine the commercial wisdom of every production, resolve uncertain rights positions or transform an inherently uncertain creative outcome into a predictable event. These questions require the combination of reliable evidence, professional expertise, legal and contractual analysis, appropriate governance and disciplined judgement.
The name UNIVERSAL INTELLIGENCE® expresses the framework's aspiration towards a whole-system perspective. Its purpose is to connect relevant dimensions of a production sufficiently well to understand how they interact, while recognising that no model or institution can possess complete knowledge. The commercial framework is concerned with the practical application of artificial intelligence to production insurance and with the development of a more coherent institutional capability for understanding creative risk.
The future of insurability
The future of production insurance will be shaped by continuing changes in the technology, economics and organisation of creative work. Production methods will evolve; the use of synthetic and digitally modified material will develop; contractual arrangements will respond to new practices; and the evidence required to assess certain exposures may change. Insurers will need to understand these developments without assuming that historical experience always provides an adequate guide or that every new capability creates an insurable opportunity.
The institutions best equipped to meet these changes will be those able to combine analytical sophistication with intellectual discipline. They will use technology to investigate complex dependencies, but they will also recognise uncertainty, question assumptions and retain accountability for consequential decisions. They will distinguish the possibility of a loss from evidence that it is likely to occur, the identification of an exposure from the determination of its insurability and a plausible model output from a conclusion that can be defended in practice.
Artificial intelligence may make more evidence accessible and enable relationships to be examined at a greater scale. Its enduring value, however, will depend upon the institution's capacity to convert that capability into sound judgement. This requires continuing research, reliable data, appropriate governance, institutional memory and a culture that is willing to revise its understanding when the evidence changes. It also requires restraint: the recognition that some uncertainties cannot be resolved and that a responsible decision may sometimes be to seek further evidence, impose conditions or decline an exposure.
The purpose of intelligence is therefore not to make every production insurable. It is to improve the quality of the decisions through which insurability is assessed. By understanding more deeply the relationships that determine how productions operate, how losses may develop and how contractual risk can be allocated, insurers may be better equipped to distinguish risks that can be accepted on defensible terms from those that cannot. The quality of that distinction is central to the long-term credibility of the insurance relationship.
The enduring proposition
Creative enterprise has always depended upon a willingness to act despite uncertainty. Films and television programmes are conceived before their final form is known, financed before their eventual reception can be established and produced through arrangements that must remain sufficiently resilient to accommodate unforeseen circumstances. Insurance contributes to this process by providing a contractual means of transferring defined risks. Its future effectiveness will depend upon the continuing quality of the judgement through which those risks are understood and accepted.
UNIVERSAL INTELLIGENCE® provides a framework for examining how artificial intelligence may contribute to that judgement. Its focus is the production as a whole: the relationships between its creative, operational, financial, contractual and technological dimensions and the conditions under which these relationships may create or amplify risk. By helping insurers connect evidence, examine assumptions, recognise dependencies and learn from experience, artificial intelligence may support a more comprehensive understanding of the undertaking. That understanding must remain proportionate to the evidence and disciplined by the limits of what can reasonably be known.
The future of insurability will not be determined by the elimination of uncertainty, because uncertainty is inseparable from the creative process itself. Nor will it be secured by technological sophistication alone. It will depend upon the capacity of insurers to distinguish what can be understood from what remains uncertain, what can be mitigated from what can be transferred and what can reasonably be accepted from what cannot be supported on defensible terms. Intelligence provides no guarantee of a favourable outcome; it provides the means of making better-informed decisions about the risks that must be assumed.
For GENERAL INTELLIGENCE PLC, the purpose of UNIVERSAL INTELLIGENCE® is to bring this principle into the practical domain of production insurance. It is concerned with the development of institutions capable of understanding complex creative undertakings more comprehensively, assessing their exposures more rigorously and exercising underwriting judgement with greater discipline. Its objective is not to replace professional expertise, but to strengthen the intellectual environment within which that expertise operates.
The enduring ambition is to make creative uncertainty more intelligible without pretending that it can be removed; to make the assessment of risk more comprehensive without obscuring the limits of knowledge; and to support the insurability of creative enterprise through better understanding rather than unjustified certainty.
The future of production insurance will depend not upon the elimination of uncertainty, but upon the quality of the intelligence with which it is understood, assessed and assumed.