Adaptation occupies a fundamental position in the history of ideas concerning intelligence, life and the relationship between an organism and its environment. To adapt is to adjust to circumstances, to accommodate changing conditions or to modify behaviour in response to new demands. The concept extends across biology, psychology, philosophy, economics and organisational theory, although its meaning differs between disciplines. In each, it raises a common question: what enables an individual, system or institution to remain effective when the conditions upon which its existing behaviour depends are no longer the same?
The English adjective adaptive emerged in the late eighteenth century, while the verb adapt derives through French from the Latin adaptare, meaning to fit or adjust. The word intelligence has a longer recorded history in English, appearing in the late fourteenth century through Old French and Latin forms associated with understanding, knowledge and discernment. The combination of the two ideas expresses a distinction between intelligence as an ability and intelligence as an ability exercised in changing circumstances. It suggests that the value of understanding lies not merely in what an individual or institution knows, but in whether that knowledge enables an appropriate response to the conditions actually encountered.
The intellectual history of adaptive intelligence cannot, however, be reduced to one invention or attributed confidently to a single originator. The underlying relationship between intelligence and adaptation precedes the modern terminology by centuries, while the explicit concept of adaptive intelligence has developed through different strands of psychological and scientific thought. Its contemporary formulation is particularly associated with the work of the psychologist Robert J. Sternberg, whose 2019 paper, A Theory of Adaptive Intelligence and Its Relation to General Intelligence, examined whether established conceptions of general cognitive ability adequately captured intelligence understood as adaptation to the environment. His subsequent book, Adaptive Intelligence: Surviving and Thriving in Times of Uncertainty, published by Cambridge University Press in 2021, developed the argument further.
Sternberg's work is an important modern contribution to the theory of adaptive intelligence, not proof that the phrase originated with him. Nor should a theory developed principally in relation to human intelligence be treated as identical to an institutional framework for insurance. The intellectual connection is more fundamental. Both concern the capacity to respond intelligently when circumstances change, existing assumptions become less reliable or familiar methods no longer provide an adequate guide to action.
For a general insurance company, this relationship has particular significance. Insurance is founded upon the assessment of uncertain future events, yet the evidence upon which that assessment depends is frequently derived from the past. Historical claims, exposure data, actuarial models and underwriting experience remain indispensable. Their usefulness, however, depends upon the continued relevance of the conditions from which they were derived. When those conditions change, the institution must determine whether its existing understanding remains adequate.
ADAPTIVE INTELLIGENCE® applies this problem to the modern insurer. It is concerned with the ability to recognise change, interpret its significance, reconsider established assumptions and translate new evidence into proportionate action. It does not imply perpetual modification, automatic responsiveness or an ability to predict every emerging risk. Its ambition is more disciplined: to develop an institutional capacity for knowing when circumstances have changed sufficiently to require a different understanding and when they have not.
From adaptation to intelligence
The idea of adaptation has played a central role in biology, where the relationship between living organisms and their environments has long been a subject of inquiry. The development of evolutionary theory, particularly through Charles Darwin's account of natural selection, established the importance of understanding how characteristics that affect survival and reproduction interact with environmental conditions over generations. Adaptation in this context is not necessarily a conscious process, nor does it imply foresight or deliberate intelligence. It describes a relationship between characteristics, changing conditions and their consequences.
The distinction is important because the term adaptive intelligence can otherwise suggest a more unified theory than the historical record supports. Biological adaptation, human cognitive adaptation, organisational adaptation and adaptive behaviour in artificial intelligence are related through a broad concern with adjustment to circumstances, but they are not equivalent processes. An organism may adapt through evolutionary mechanisms without consciously learning; a person may adjust a decision in response to experience; an algorithm may update its parameters under a defined learning procedure; and an institution may revise its practices through governance and professional judgement. Each involves a relationship between existing capability and changing conditions, but the mechanisms and standards of success differ.
In psychology, the concept of intelligence has often been approached through the measurement of cognitive abilities. Charles Spearman's work on general intelligence, published in 1904, was influential in establishing the investigation of common factors underlying performance across different mental tasks. Subsequent research developed a range of models, none of which should be confused with a complete account of intelligent behaviour in every real-world setting. The capacity to perform well on established cognitive tasks does not necessarily establish the ability to respond wisely to unfamiliar circumstances, account for wider consequences or distinguish short-term success from longer-term adaptation.
Sternberg's modern account of adaptive intelligence draws attention to precisely this distinction. It asks whether intelligence should be evaluated solely through traditional measures of cognitive ability or also by considering how thought and behaviour function within biological, cultural and practical environments. The argument does not require the rejection of general intelligence research. It invites a wider examination of what intelligence is intended to explain and whether conventional measures adequately capture the forms of judgement that matter in changing circumstances.
The relevance to institutions follows naturally. An organisation may possess sophisticated models, extensive historical records and highly capable professionals while remaining poorly equipped to respond when its environment changes. Its systems may continue to produce consistent outputs, even as the assumptions underlying those outputs lose validity. Its established procedures may be efficient under familiar conditions but fail to recognise a new form of exposure. Its accumulated expertise may be valuable, yet overly dependent upon experiences that no longer represent the risks it faces.
Institutional adaptation requires more than the ability to change. It requires the intelligence to establish when change matters, why it matters, what should be altered and which existing capabilities remain sound. Without this distinction, adaptation can become reactivity: a succession of adjustments driven by isolated events rather than a coherent understanding of the environment. Adaptive intelligence seeks the opposite. It is directed towards continuity of purpose through informed reconsideration of means.
Insurance and the limits of historical knowledge
General insurance is particularly exposed to the tension between historical experience and changing circumstances. The industry must estimate uncertain future losses, establish premiums and terms, maintain appropriate reserves, manage capital and preserve the ability to meet obligations to policyholders. These functions depend upon evidence, statistical methods, actuarial expertise, underwriting judgement and an understanding of the environment in which risks arise.
Historical data remains essential to this work. Claims records can reveal the frequency and severity of losses; exposure information can help establish the characteristics of insured risks; and actuarial analysis can identify relationships that inform pricing and reserving. Yet no historical dataset can guarantee that the future will reproduce the conditions from which it was assembled. A change in technology, economic conditions, legal interpretation, customer behaviour or the physical environment may alter the distribution of losses. The importance of historical evidence lies not in its presumed permanence, but in its capacity to inform an assessment whose relevance must be continually examined.
The problem becomes particularly acute where the environment is changing faster than conventional analytical processes can incorporate new evidence. Inflation may alter repair and replacement costs. Emerging technologies may create liabilities for which claims experience is limited. Cyber risks may evolve as systems and methods of attack change. Legal and regulatory developments may affect the interpretation of existing exposures. Climate-related changes may alter the frequency, severity or geographical distribution of certain losses, while demographic, economic and behavioural developments may affect other classes of business in different ways. These are not one common risk and they should not be treated as though they share a single cause or method of prediction. Their common significance is that they may change the relationship between past evidence and future outcomes.
An insurer therefore requires more than a reliable account of what has happened. It requires a means of determining whether the conditions that produced those outcomes remain relevant. A model may continue to calculate correctly according to its specification while becoming less useful as a guide to a changed environment. A portfolio may appear adequately diversified under historical relationships while becoming vulnerable to a new common dependency. A claims trend may initially seem exceptional but subsequently reveal a structural development. Conversely, a conspicuous deviation may prove temporary and justify no material change in the insurer's approach.
The central challenge is interpretation. The insurer must distinguish genuine change from ordinary variation, identify the evidence supporting that judgement and determine whether the implications are sufficiently important to warrant action. ADAPTIVE INTELLIGENCE® is concerned with developing this capacity throughout the institution, rather than treating adaptation as a periodic strategic exercise or an isolated response to adverse experience.
The adaptive insurer
An insurer can be understood as an institution that continuously receives evidence about the risks it accepts. Proposals, policy information, claims, loss assessments, risk surveys, market conditions, economic indicators and external developments each reveal something about the environment in which the business operates. These sources differ in quality, timing and relevance and their significance may depend upon how they are interpreted together. The challenge is not simply to collect more information, but to establish a coherent relationship between observation, analysis, judgement and action.
An adaptive insurer possesses the capacity to use this evidence to reconsider its understanding of risk. It can identify where a change in the environment may affect underwriting assumptions, determine whether a claims pattern warrants further investigation, examine whether exposures are accumulating in a different way and assess whether an established model remains appropriate. It can distinguish the need for immediate intervention from the need for monitoring, further evidence or no change at all.
The process may be expressed as a continuing cycle: observe, interpret, challenge, decide, act and learn. Each stage is necessary. Observation without interpretation produces information without sufficient understanding. Interpretation without challenge can reinforce existing assumptions. Decisions without clear authority can create uncertainty about responsibility. Action without subsequent evaluation provides limited evidence about whether the intervention was appropriate. Learning without disciplined review can confuse coincidence with causation.
Artificial intelligence can help strengthen the connections between these stages. It can examine structured and unstructured information, identify anomalies, compare developments over time, surface contradictory evidence and help connect findings from underwriting, claims, risk management and portfolio analysis. It may allow an insurer to monitor a larger range of conditions more continuously than conventional manual processes permit. Its strategic value, however, depends upon whether the institution can interpret the resulting signals and translate them into decisions that are justified by the evidence.
The adaptive insurer is not one that changes its position whenever a new observation appears. Such behaviour would undermine consistency and could make underwriting more volatile rather than more intelligent. It is one that understands which changes matter, which assumptions they affect and what response is proportionate. Adaptation is therefore governed by judgement. The objective is a shorter and more reliable connection between material change in the real world and the institution's recognition of its significance.
Underwriting as a continuing intelligence process
Underwriting is traditionally organised around a defined decision. Information is gathered, the exposure is assessed, the terms are determined and the risk is accepted or declined. This process remains necessary. The insurer must establish what it is being asked to insure, evaluate the evidence and determine the conditions upon which it is prepared to assume responsibility. ADAPTIVE INTELLIGENCE® does not seek to weaken that discipline or substitute continuous machine activity for professional underwriting.
Its contribution lies in recognising that the environment surrounding a risk may continue to evolve after the original decision. The significance of this evolution depends upon the class of business, the policy, the nature of the exposure and the relevant contractual arrangements. In some circumstances, a change may have no material consequence for the risk accepted. In others, new information may warrant reconsideration of the insurer's understanding, its exposure management or its approach at renewal. The appropriate response must always be consistent with applicable law, policy terms and the responsibilities owed to policyholders.
Artificial intelligence can assist by monitoring relevant developments, comparing current information with the assumptions underlying earlier decisions and identifying changes that warrant professional attention. It may help underwriters distinguish an isolated anomaly from a potentially broader shift, examine whether historical comparators remain appropriate and identify where a risk assessment depends upon information that has become outdated or incomplete.
This creates a more continuous relationship between underwriting and the risk environment. The underwriter's responsibility remains to assess the evidence and determine its significance. The intelligence system supports that responsibility by making relevant information more accessible and by helping identify when the basis for the original judgement may deserve review.
The distinction between adaptation and automatic repricing is essential. Continuous monitoring does not itself confer authority to change policy terms or alter contractual obligations. Nor does new information necessarily justify an intervention. The purpose is to improve the insurer's awareness and understanding, allowing appropriate decisions to be made through established governance and contractual processes. Intelligence should strengthen the quality of action, not obscure the rules under which action is permissible.
Claims as evidence of changing risk
Claims provide one of the most direct sources of information about the relationship between insured risks and actual outcomes. They reveal the circumstances in which losses occur, the extent of the resulting damage, the costs of repair or settlement, the time required to resolve claims and the effectiveness of arrangements intended to reduce harm. Across a portfolio, claims experience may also reveal patterns that are difficult to identify from individual cases alone.
An adaptive insurer treats this experience as a continuing source of intelligence. A claim is not merely a financial obligation to be managed under the policy; it is also evidence that may inform future underwriting, reserving, risk management and product design. The relationship is not automatic. A single claim rarely establishes a general trend and differences between insureds, policy terms, jurisdictions and circumstances can make direct comparisons misleading. Nevertheless, carefully interpreted claims experience can reveal where assumptions require examination or where an exposure is changing in character.
Artificial intelligence can assist in connecting claims information with relevant policy, underwriting, economic and operational evidence. It may help identify recurring causes, unusual changes in claims development, inconsistencies in reserve assumptions or patterns suggesting that previously separate exposures share a common influence. The resulting analysis can direct expert attention towards matters that deserve further examination, while preserving the distinction between a signal and a substantiated conclusion.
The wider opportunity concerns learning and prevention. Where evidence indicates that particular conditions are associated with increased loss, an insurer may be able to support improved risk information, encourage proportionate mitigation or revise its assessment methods. The capacity to reduce the likelihood or severity of loss can benefit policyholders and the insurer, although the feasibility and effectiveness of prevention will differ substantially by class of business and circumstance. Not every risk can be prevented and not every observed pattern supports intervention.
The adaptive insurer therefore seeks to turn claims experience into a disciplined feedback process. It learns from what has happened, but does not assume that historical patterns will persist unchanged. It uses evidence to improve the assessment of future exposures and remains attentive to circumstances in which the lessons of the past have become less applicable.
The adaptive portfolio
The insurance portfolio represents a different level of complexity from the individual policy. Its performance depends upon the interaction of exposures, the relationship between risks, the concentration of capital and the possibility that adverse conditions may affect many policyholders simultaneously. Diversification is essential to insurance, but diversification cannot be assessed entirely by counting policies or separating them into conventional categories. Risks that appear distinct may depend upon common economic, technological, geographical or operational conditions.
Portfolio intelligence must therefore examine the relationships that determine how exposures may behave together. A change in repair costs may affect several classes of business; an external event may affect many policyholders within a particular region; a common technology dependency may generate correlated losses across otherwise unrelated insureds. The importance of such connections will vary, but their identification can be material to underwriting strategy, accumulation management, capital allocation and resilience.
Artificial intelligence can help analyse heterogeneous information across portfolios, identify potential concentrations and investigate changes in the relationships between exposures. It may allow the insurer to examine more scenarios, compare developments across classes of business and direct professional attention towards emerging patterns. These findings require careful interpretation. Correlations may be temporary, apparent similarities may conceal material differences and models based on past experience may be least reliable when the environment departs most sharply from historical conditions.
The purpose of adaptive portfolio intelligence is not to provide a machine with autonomous authority over risk appetite or capital allocation. It is to improve the institution's ability to recognise where its aggregate exposure is changing, understand the possible consequences and determine whether its existing assumptions remain defensible. This requires the integration of analytical capability with actuarial expertise, underwriting judgement, capital management and appropriate governance.
The portfolio becomes adaptive when the institution can reconsider its aggregate position in response to credible evidence, without allowing every new signal to dictate a change in strategy. Its capacity rests upon the quality of the intelligence available and the discipline with which that intelligence is translated into decisions.
Intelligence that examines itself
One of the most important dimensions of adaptive intelligence is the ability to recognise when the organisation's own methods may no longer be reliable. Insurers rely upon models, data, assumptions and professional practices to interpret risk. These tools are useful precisely because they impose structure upon uncertainty. Yet that structure may become less representative as circumstances change. Data may become incomplete or less relevant; relationships between variables may weaken; customer behaviour may evolve; and an analytical method that performed well under familiar conditions may become less effective in a different environment.
An insurer must therefore examine not only the risks it underwrites, but the continuing validity of the intelligence through which those risks are understood. A system that produces precise outputs while becoming progressively less relevant can create an especially serious form of institutional vulnerability. The appearance of consistency may conceal a deterioration in the underlying basis of the analysis.
Artificial intelligence can assist with the monitoring of data quality, model performance, anomalous outputs and changes in the characteristics of incoming information. It may help identify circumstances in which established relationships require reassessment or where the evidence supporting a prediction has become less reliable. But these capabilities also require scrutiny. A model intended to identify model deterioration may itself be limited by incomplete evidence or unsuitable assumptions. No technical system should be presumed capable of validating itself conclusively.
The governance of adaptive intelligence must consequently include independent review, defined responsibilities, appropriate validation and the ability to suspend or modify a system when its reliability cannot be established. Human expertise remains necessary to determine whether a detected change has substantive meaning and whether the response is proportionate to the consequences of error.
Adaptability, properly understood, is not a licence for unrestricted self-modification. It is the disciplined capacity to recognise when existing methods require reconsideration and to manage that reconsideration responsibly. The intelligent insurer is prepared to change, but it does not change without cause, authority or an understanding of the consequences.
Artificial intelligence as an institutional capability
The strategic significance of artificial intelligence lies in the possibility of connecting activities that have historically operated through separate systems, processes and professional responsibilities. Underwriting generates information relevant to portfolio management. Claims reveal evidence about the risks accepted. Actuarial analysis informs pricing and reserving. External information helps explain changes in the environment. Risk management examines concentrations and vulnerabilities. Each discipline contributes a different perspective, but institutional intelligence depends upon the ability to bring these perspectives together when a decision requires them.
Artificial intelligence can help make this integration more practicable. It can search across large bodies of records, compare information from different sources, identify inconsistencies and support the retrieval of relevant knowledge. It can assist professionals in investigating how a development in one part of the insurer may affect another, while helping preserve the reasoning behind consequential decisions. Its purpose is not to eliminate specialist expertise, but to make that expertise more accessible and effective within the institution as a whole.
The benefits depend upon sound institutional design. Fragmented or unreliable data cannot be made dependable merely by processing it with a more sophisticated model. Poorly defined responsibilities can become more difficult to manage when automated recommendations enter established workflows. A proliferation of separate systems may increase complexity without improving the coherence of decisions. The insurer must therefore consider data architecture, security, system integration, governance, staff capability and the relationship between analytical outputs and professional authority as part of the same strategic question.
This is also why artificial intelligence should not be treated as an end in itself. In some circumstances, a more effective process, improved information quality or better-defined professional responsibilities may create more value than advanced modelling. In others, artificial intelligence may materially improve the institution's ability to detect changes, examine interactions and learn. The appropriate choice depends upon the problem and the evidence of contribution.
A genuinely adaptive institution is not identified by the quantity of technology it deploys. It is identified by its capacity to select the right means, recognise their limitations, evaluate their results and change its approach when circumstances require it. Artificial intelligence becomes an institutional capability when it strengthens that discipline rather than merely adding another layer of technology.
Human judgement, governance and responsibility
Insurance decisions frequently require judgement under incomplete evidence. The interpretation of an unusual claim, the assessment of a novel exposure, the evaluation of a complex liability or the treatment of an exceptional circumstance may involve questions that cannot be resolved adequately through pattern recognition alone. Even where a model provides a reliable estimate, the institution must determine what that estimate means within the particular legal, contractual and commercial context.
Human intelligence is therefore central to adaptive insurance. Underwriters, actuaries, claims professionals, risk managers and executives bring experience, professional standards, contextual understanding and accountability to decisions. Artificial intelligence can extend their ability to examine information, identify patterns and challenge established beliefs. It can also introduce new risks through flawed outputs, inadequate data, inappropriate assumptions, automation bias and the difficulty of understanding how complex systems reach conclusions.
The appropriate relationship is one of disciplined cooperation. Machine systems should support the gathering and interpretation of evidence within clearly defined boundaries; human professionals must retain the authority and responsibility appropriate to the decision. The greater the potential consequence of error, the more important it becomes to establish the reliability of the system, preserve the ability to challenge its outputs and ensure that responsibility cannot become obscured by automation.
Governance is not separate from intelligence. An institution that cannot explain how a consequential decision was reached, identify who authorised it or correct a materially flawed output lacks an essential form of institutional understanding. Adaptive intelligence therefore requires appropriate oversight, validation, monitoring, records, security and accountability. It also requires the ability to recognise uncertainty and refrain from acting where the evidence does not justify a confident conclusion.
This principle is consistent with the practical use of artificial intelligence within insurance supervision. In December 2025, the Bank of England described the use of AI-assisted pattern detection to analyse general insurance reserving data, with the objective of identifying potentially significant anomalies more efficiently while keeping human supervisors involved in interpretation and challenge. The example illustrates a relevant distinction: technology can widen the field of observation and direct professional attention, but the quality of judgement and oversight remains essential to the value of the result.
For general insurance companies, the lesson is not that every decision should be automated or that supervisory methods can be transferred unchanged into commercial underwriting. It is that greater analytical capability is most useful when it improves the focus, consistency and timeliness of professional judgement while preserving institutional responsibility.
ADAPTIVE INTELLIGENCE® at GENERAL INTELLIGENCE PLC
ADAPTIVE INTELLIGENCE® is an artificial intelligence consultancy framework provided by X from GENERAL INTELLIGENCE PLC for general insurance companies. It addresses the strategic, technical and institutional dimensions of intelligence within an industry whose effectiveness depends upon understanding risk in an environment that is continually changing.
The framework considers how artificial intelligence may help insurers perceive developments in their risk environment, interpret evidence across underwriting, claims and portfolio management, test the continued validity of established assumptions and improve the relationship between information and action. Its purpose is to develop a more coherent institutional capability for responding to material change, rather than treating artificial intelligence as a collection of isolated technical applications.
The framework begins with the insurer's actual requirements. It asks where the institution's understanding is constrained, which changes in the environment may affect its decisions and whether artificial intelligence can make a demonstrable contribution. It examines the relationship between historical evidence and current conditions, the recognition of emerging patterns, the monitoring of portfolio exposure, the intelligence derived from claims and the processes through which the organisation revises its understanding.
It also considers the limits of technological adaptation. Predictive capability is not equivalent to reliable foresight; changing a model is not the same as changing an institution intelligently; and the rapid production of analytical outputs does not guarantee that those outputs are material or correct. The framework therefore places emphasis upon professional judgement, the quality of evidence, appropriate validation, proportionality and clear accountability. Its purpose is not to maximise the speed of change, but to improve the quality of the decisions through which change is interpreted and managed.
ADAPTIVE INTELLIGENCE® is concerned with the insurer as a learning institution. Underwriting, claims, actuarial analysis, portfolio management, risk governance and external intelligence should be capable of informing one another where relevant. The institution should be able to preserve the reasoning behind important decisions, identify where assumptions require reconsideration and learn from experience without treating historical patterns as permanent truths. Artificial intelligence may support this process, but its success depends upon the organisation's ability to integrate the resulting knowledge into established responsibilities and decision-making.
The designation ADAPTIVE INTELLIGENCE® expresses the particular consultancy framework developed for this purpose. The phrase itself belongs to a wider intellectual discussion of intelligence and adaptation and its history should not be represented as originating with one organisation. GENERAL INTELLIGENCE PLC's use of the designation identifies its defined commercial offering for general insurance companies. The framework's significance lies in its specific application to the changing nature of insured risk and the institutional capability required to understand that change responsibly.
The adaptive intelligence test
The test of ADAPTIVE INTELLIGENCE® is whether an insurer can recognise material change earlier, establish its significance more reliably and respond more appropriately than it could before.
- Can it distinguish ordinary variation from a genuine shift in risk?
- Can it identify when historical evidence has become a less reliable guide?
- Can it connect claims experience with underwriting decisions and portfolio management?
- Can it recognise emerging concentrations without mistaking superficial similarities for common exposure?
- Can it determine when a model deserves less confidence?
- Can it adapt without becoming reactive?
The test also concerns institutional discipline.
- Can the insurer preserve independent professional judgement as artificial intelligence becomes more capable?
- Can it establish which decisions may be automated, which require human authority and how errors can be identified and corrected?
- Can it learn from experience without allowing hindsight to distort the evaluation of past decisions?
- Can it use new technology while maintaining appropriate security, governance and accountability?
- Can it distinguish the introduction of technical capability from the creation of actual institutional value?
These questions are more demanding than a measure of processing speed or the number of tasks automated. They assess whether the institution has improved the relationship between evidence, understanding and action. A system that detects every minor deviation but cannot distinguish material developments from noise may add complexity rather than intelligence. A model that adapts continuously without sufficient control may become unstable. An insurer that reacts to every emerging signal may sacrifice consistency without improving risk selection.
The objective is disciplined responsiveness. Adaptation is valuable when it reflects a credible understanding of changed circumstances, an assessment of the resulting implications and a proportionate response. It is not an obligation to change for the sake of change, nor a promise that every risk can be anticipated. It is the institutional capacity to reconsider an established position when the evidence warrants it, while retaining the judgement necessary to determine when the existing position remains justified.
The enduring proposition
The history of adaptive intelligence reflects a broader intellectual concern with the relationship between capability and circumstance. The idea that intelligence should be considered in relation to adaptation has developed through psychological theory and wider discussions of cognition, learning and the ability to respond to an environment. Its application to insurance gives that idea a distinct institutional meaning. An insurer must make decisions about future losses using evidence that is necessarily incomplete, often historical and sometimes least reliable when conditions are changing most rapidly.
ADAPTIVE INTELLIGENCE® applies this concern to the strategic, technical and institutional challenges facing general insurance companies. It considers how artificial intelligence may improve the capacity to perceive change, interpret emerging evidence, challenge assumptions, connect information across the enterprise and convert experience into institutional learning. Its purpose is not to remove uncertainty from insurance, but to improve the quality of the decisions made in response to it.
The future of general insurance will not be determined by the quantity of data collected, the sophistication of individual models or the speed with which every process can be automated. It will depend, in significant part, upon whether institutions can understand when their environment has changed, establish what that change means for the risks they accept and act with appropriate discipline. That requires technology, professional expertise, reliable evidence, sound governance and a willingness to revise conclusions when circumstances demand it.
The adaptive insurer is therefore not an organisation in constant motion. It is an institution capable of maintaining its purpose while reconsidering its assumptions; of learning from experience without becoming imprisoned by it; and of using artificial intelligence to extend its understanding without surrendering responsibility for its decisions. Its distinguishing quality is not an ability to predict everything, but the capacity to recognise when what it knows is no longer sufficient.
For GENERAL INTELLIGENCE PLC, this is the central proposition of ADAPTIVE INTELLIGENCE®: that the enduring value of artificial intelligence lies in the development of institutions capable of understanding a changing world, learning from the evidence it provides and responding with disciplined judgement.