As artificial intelligence becomes increasingly accessible across the insurance market, managing general agents face a question more consequential than which technologies to adopt: how can they extend the intelligence that makes them valuable without losing the specialist expertise on which their competitive position depends?
X from GENERAL INTELLIGENCE PLC has developed BIONIC INTELLIGENCE®, an artificial intelligence consultancy framework specifically for managing general agents (MGAs). Its central proposition is that the strongest MGA of the future will not simply automate an existing underwriting business. It will deliberately integrate human underwriting intelligence with machine intelligence, advanced analytics and institutional knowledge to create a more capable, connected and continuously learning organisation.
The distinction matters. An MGA's competitive advantage often resides in a relatively small number of experienced professionals who understand a particular class of business, recognise subtle differences between risks, maintain valuable broker relationships and exercise judgement that cannot readily be reduced to a standardised model. Their expertise is commercially valuable precisely because it is specialised. Yet even the most accomplished underwriter has finite attention, cannot continuously examine every relevant development and cannot independently maintain a complete understanding of every relationship within an expanding portfolio.
BIONIC INTELLIGENCE® addresses this tension between the depth of specialist human judgement and the breadth of analysis increasingly available through machines. Rather than treating the underwriter as an obstacle to automation, it asks how the organisation can extend that professional's capacity to perceive, investigate, compare, remember and decide.
The objective is not to make the underwriter obsolete. It is to make exceptional underwriting capability more scalable, more consistently informed and more enduring.
From technological adoption to underwriting advantage
For an MGA, the commercial test of artificial intelligence is not whether it can perform a task more quickly. It is whether the resulting capability improves the quality of underwriting decisions and creates value that can be sustained.
Administrative automation may reduce friction, but efficiency alone does not necessarily strengthen risk selection, improve pricing or distinguish one underwriting business from another. The more consequential opportunity lies in using machine intelligence to connect information that is otherwise dispersed across submissions, policies, claims, broker relationships, market developments and the MGA's own experience.
Such connections may help underwriters identify unusual combinations of risk, investigate emerging exposures, test assumptions and recognise changes that deserve attention. They may also reveal opportunities that would remain difficult to identify through isolated reviews or conventional reporting.
The professional remains responsible for interpreting the significance of the evidence. Artificial intelligence can broaden the field of analysis; it does not, by itself, establish what a particular risk means, what price is appropriate or whether an opportunity fits the MGA's commercial purpose.
The portfolio behind the individual risk
The framework places particular emphasis on the relationship between individual underwriting decisions and the aggregate portfolio they create.
An individual risk may appear attractive when assessed in isolation, while its relationship with other accepted risks creates a concentration that is less apparent. Geographical accumulation, common dependencies, correlated exposures and changes in business mix may emerge gradually across a portfolio assembled through many separate decisions.
BIONIC INTELLIGENCE® treats portfolio intelligence as a continuous discipline. Artificial intelligence and analytical systems may help identify patterns across risks, compare actual business with intended appetite and highlight changes that warrant investigation. The purpose is to make the portfolio more intelligible to the organisation responsible for it—not to substitute an algorithmic assessment for informed underwriting judgement.
This has particular relevance to delegated authority. An MGA operates within authority entrusted to it by another institution. Understanding whether individual transactions remain within specified parameters is important; understanding whether the portfolio as a whole continues to reflect the economic and risk assumptions behind that authority is equally significant.
The framework therefore treats delegated authority not simply as a set of permissions and compliance requirements, but as an intelligence architecture connecting underwriting decisions, portfolio development, performance monitoring and institutional responsibility.
Making every claim an opportunity to learn
The relationship between underwriting and claims provides another important dimension.
A claim can reveal more than the circumstances of an individual loss. It may expose an assumption that was incomplete, a weakness in pricing, an ambiguity in policy wording, a gap in risk selection or a change in the environment in which the MGA operates.
Where underwriting and claims information remain separated, the lessons may arrive late or fail to influence subsequent decisions. BIONIC INTELLIGENCE® seeks to connect the original underwriting rationale with subsequent experience, enabling the organisation to investigate what happened, what its earlier assessment understood or missed and whether the evidence warrants a change in appetite, pricing, wording or selection.
The ambition is to shorten the distance between experience and learning. This does not mean that every claim produces a definitive lesson, or that patterns identified by analytical systems are necessarily causal. It means establishing a more systematic way to examine experience and allow substantiated findings to inform future decisions.
The same principle applies to exceptions, unusual submissions, referrals and unexpected market developments. Rather than treating every departure from an established pattern as an administrative inconvenience, the MGA can investigate whether it represents a source of risk, an emerging opportunity or evidence that its existing understanding requires revision.
Human expertise as a strategic asset
BIONIC INTELLIGENCE® also addresses a problem that can be particularly acute in specialist underwriting businesses: the fragility of institutional memory.
An MGA may depend upon a small number of individuals whose understanding has developed over many years. Their knowledge may encompass the history of a market, the behaviour of particular risks, the reasons behind previous underwriting decisions and the lessons of claims that never became formal guidance. Much of this expertise may remain tacit, distributed across personal experience, working relationships and disconnected records.
Artificial intelligence may help preserve and retrieve the history of decisions, referrals, exceptions, claims, broker interactions and market observations. Properly designed, such capabilities can make accumulated knowledge more accessible to the next generation of underwriters without pretending that experience can be captured completely in a database.
The institutional objective is continuity: enabling the organisation to retain more of what it has learned, understand why decisions were made and give professionals better access to the evidence and reasoning that preceded them.
This is also a question of organisational design. Information should flow between submissions, underwriting, pricing, binding, claims, portfolio management, brokers and capacity providers. Each function should be capable of informing the others, so that decisions generate evidence, evidence produces learning and learning can improve subsequent decisions.
Agility requires governance
The MGA model is associated with specialist expertise, entrepreneurial judgement and the ability to respond to market opportunities. The use of increasingly capable artificial intelligence should strengthen these characteristics, not introduce an opaque layer of decision-making that makes the business harder to understand or control.
Delegated underwriting authority makes transparency, data quality, validation, monitoring, auditability and clear human responsibility especially important. An organisation should understand not only what a system recommends, but the evidence and assumptions behind the recommendation, the limits of its reliability and the circumstances in which a professional must intervene.
BIONIC INTELLIGENCE® consequently rejects the proposition that a smaller, more entrepreneurial institution should accept weaker governance because it is more agile. Properly designed controls can support responsible speed by making the boundaries of decision-making clearer.
The framework is also deliberately technology-neutral. Depending on the objective, better data, clearer underwriting authority, improved training, a redesigned process or stronger organisational discipline may offer greater value than a new artificial intelligence application. The technology must justify its place through its contribution to the MGA's purpose.
A different conception of the MGA
BIONIC INTELLIGENCE® forms part of the wider research and consultancy work of X from GENERAL INTELLIGENCE PLC concerning human intelligence, artificial intelligence, machine intelligence and institutional intelligence.
Its proposition is that the future competitive distinction between MGAs may depend not simply on which organisations possess advanced technology, but on which can establish the most effective relationship between specialist human expertise and machine capability.
That relationship cannot be created by technology alone. It requires decisions about what the organisation values, what its underwriters need to understand, how authority should be exercised, how evidence should be evaluated and how lessons should be retained. It requires professional judgement and institutional accountability to remain explicit as analytical capacity expands.
The result, if successfully designed and implemented, would be an MGA better equipped to examine its risks, understand its portfolio, allocate professional attention and learn from experience. The precise benefits would depend upon the organisation, the quality of its information, the suitability of the chosen methods and the discipline with which they are governed.
X from GENERAL INTELLIGENCE PLC's starting principle is straightforward: artificial intelligence should be assessed by the value it contributes, not by its novelty or sophistication.
Artificial intelligence is the means. BIONIC INTELLIGENCE® is the capability. Underwriting judgement is the discipline. Sustainable underwriting advantage is the objective.
FURTHER INFORMATION:
1. About BIONIC INTELLIGENCE®
BIONIC INTELLIGENCE® is an artificial intelligence consultancy framework provided by X from GENERAL INTELLIGENCE PLC to managing general agents.
It examines how an MGA can integrate specialist human underwriting judgement with machine intelligence, advanced analytics, institutional data and intelligent operating processes. Its central concern is the design of an organisation in which technology extends professional capability rather than simply automating existing tasks.
The framework addresses underwriting, delegated authority, binding authority, risk appetite, selection, pricing, referral, exceptions, broker intelligence, claims, capacity-provider relationships, portfolio analysis, growth, market change, institutional memory and governance.
2. What is a managing general agent?
A managing general agent is an insurance intermediary or specialist underwriting business entrusted with defined authority by an insurer or other capacity provider. Depending on its arrangements, an MGA may perform functions including underwriting, binding risks, administering policies and managing aspects of the insurance lifecycle.
MGAs can occupy a distinctive position in the insurance market. They may combine specialist underwriting knowledge and distribution relationships with access to capacity provided by another institution. Their success can depend on understanding a particular class of business more deeply than a generalist competitor, while maintaining appropriate discipline over the risks they accept.
The specific activities and responsibilities of an MGA depend on its contractual arrangements, permissions and operating model.
3. Why the framework uses the term “bionic”
The word bionic describes the deliberate combination of biological capability and engineered capability. In this framework, it refers to the integration of human and machine intelligence within an underwriting institution.
The concept is one of augmentation rather than substitution. Machine systems may contribute scale, speed, retrieval, comparison, pattern recognition and continuous monitoring. Human professionals contribute contextual reasoning, commercial judgement, interpretation, negotiation, ethical consideration and accountability.
The purpose is to design the relationship between these capabilities so that each contributes where it is best suited.
4. The central commercial question: does intelligence improve underwriting?
The framework distinguishes the adoption of technology from the creation of underwriting advantage.
A system may make an existing process faster without materially improving the quality of the underlying decision. A more consequential application may help an MGA identify a previously obscured exposure, examine a complex submission, recognise an emerging pattern or understand how individual risks interact within the portfolio.
The appropriate test is whether the capability improves the organisation's ability to understand and decide—not merely whether the technology performs an impressive technical function.
The value of any application must be evaluated in its specific commercial and operational context.
5. The bionic underwriter
The bionic underwriter is a professional whose effective field of analysis is extended by access to relevant information and analytical capability.
Potential sources of information include:
- Current and historical submissions.
- Policy and claims records.
- Market and specialist information.
- Publicly available information, where relevant and lawfully used.
- Geospatial and economic information.
- Broker activity and distribution patterns.
- The MGA's own underwriting decisions, referrals and accumulated experience.
Artificial intelligence may help summarise, compare, classify, identify anomalies, test assumptions and surface relationships across these sources. The underwriter remains responsible for evaluating their relevance, challenging recommendations and reaching the professional decision.
These are potential applications of the framework, not claims that every MGA possesses the necessary data or that each application will be appropriate in every setting.
6. Delegated authority as an intelligence architecture
Delegated authority is central to the MGA model. An MGA's authority is granted within defined parameters and its exercise has implications for both the MGA and the institution providing capacity.
The framework considers how analytical systems may help an MGA understand:
- What business it is writing and where.
- Whether actual business remains consistent with underwriting appetite.
- How individual decisions aggregate across the portfolio.
- Where concentrations or deviations are developing.
- Whether changes in business mix affect the assumptions behind the delegation.
- Which transactions or patterns warrant investigation or referral.
This approach goes beyond checking whether a transaction complies with a rule. It asks whether the business being written continues to reflect the strategic and economic purpose for which authority was granted.
Any such monitoring must be aligned with the relevant authority agreements, data arrangements and governance responsibilities.
7. Binding authority and the accumulation of risk
Binding authority can make specialist underwriting repeatable and scalable. It can also allow concentrations to develop when each risk is assessed separately and the aggregate effect is not sufficiently visible.
The framework considers how portfolio analysis may reveal relationships across geography, industry, ownership, counterparties, supply chains, weather exposure, economic conditions and other dependencies.
The aim is to help the MGA examine the portfolio as a connected whole. Analytical signals should inform investigation and judgement; apparent correlations must not automatically be treated as evidence of causation or as definitive measures of exposure.
8. Risk appetite as a living institutional proposition
Underwriting appetite expresses the kinds of risk an MGA intends to accept and the commercial rationale for doing so.
The framework treats appetite as more than a static document consulted when a difficult case arises. It considers how actual business, emerging market conditions, changes in risk quality and the development of the portfolio can be compared with the MGA's intended position.
This can help distinguish growth consistent with underwriting strategy from growth that appears attractive in volume or premium terms but may weaken risk quality.
The objective is not to replace the professional formulation of appetite with an algorithm. It is to make the relationship between appetite and actual underwriting more visible and subject to informed review.
9. Risk selection and pricing
Risk selection and pricing are closely related but distinct professional disciplines.
Selection concerns whether a risk should be accepted and how it fits the MGA's appetite and portfolio. Pricing concerns the financial terms on which that risk should be written, taking account of expected loss, uncertainty, capital, competition, market conditions and the MGA's objectives.
Artificial intelligence may support analysis across these dimensions, identify combinations that warrant closer examination and highlight changes that may justify a review.
The framework does not advocate opaque or unchallengeable pricing. A professional should be able to examine the evidence behind a recommendation, understand its assumptions and consider the commercial circumstances before deciding whether it is appropriate.
10. Referral and the allocation of human attention
Experienced underwriting attention is a scarce resource. Routine cases, unusual submissions and potentially consequential risks do not necessarily warrant identical levels of scrutiny.
The framework considers how analytical systems may help direct straightforward cases through appropriate processes while bringing complex, unusual, high-value or potentially consequential cases to the attention of experienced professionals.
Referral is therefore treated as an allocation of expertise rather than simply an administrative stage.
The objective is to direct professional judgement where it is likely to contribute most, while retaining suitable safeguards against missed signals, inappropriate classification and over-reliance on automated recommendations.
11. Exceptions as sources of intelligence
Exceptions include unusual submissions, unexpected claims, atypical policy structures and deviations from established patterns.
An exception may indicate a problem, but it may also reveal an emerging opportunity, a change in market conditions or a limitation in the MGA's existing assumptions.
The framework encourages organisations to examine exceptions systematically. Machine intelligence may help identify deviations at scale; experienced professionals must determine which are significant and what, if anything, should follow.
Over time, a properly governed process can help an MGA build a more useful understanding of unusual cases and the circumstances in which they matter.
12. Broker intelligence
Broker relationships are central to many MGA business models. Brokers may hold information about an insured, the circumstances of a submission, competitive conditions and the commercial context of a risk that is not fully represented in structured datasets.
The framework considers how submission history, broker activity, conversion, retention and changes in business mix might be analysed alongside underwriting outcomes.
The purpose is to improve the MGA's understanding of the information available through its distribution relationships—not to mechanise those relationships or reduce brokers to numerical profiles.
Any use of broker-related data should respect its provenance, quality, lawful use and the commercial context in which it was obtained.
13. Claims as a feedback mechanism
Claims provide evidence about the risks an MGA has chosen to insure and the assumptions underpinning those decisions.
The framework considers how claims experience can be connected with the original underwriting rationale, pricing, policy wording, selection criteria and appetite. Such connections may help identify questions that merit further investigation and bring relevant experience into subsequent decisions.
A claim does not automatically demonstrate that the original underwriting decision was wrong. Outcomes are affected by uncertainty and circumstances that may not have been foreseeable at the time.
The discipline is to learn from claims without confusing hindsight with evidence that a different decision would necessarily have been correct.
14. Capacity-provider relationships
The relationship between an MGA and its capacity provider is fundamental to the business model.
Capacity providers entrust underwriting authority and capital on the basis of the MGA's specialist expertise, distribution, controls and ability to manage the risks it accepts. Portfolio visibility, transparent evidence, performance monitoring and the timely identification of emerging issues can be relevant to the quality of that relationship.
BIONIC INTELLIGENCE® considers how stronger institutional intelligence may help an MGA explain not only what it has written, but why it has written it, how the portfolio is developing and what the organisation is learning.
The framework does not imply that analytical systems alone establish the adequacy of an MGA's performance or replace contractual reporting and oversight.
15. Growth as an intelligence problem
Premium growth is not, by itself, proof of improved underwriting quality or sustainable value.
Growth may reflect genuine specialist advantage, favourable market conditions, expansion into unfamiliar business, changes in pricing or a weakening of selection discipline. The framework therefore considers how new business can be evaluated against the existing portfolio, appetite, emerging claims experience and the economic characteristics of the MGA.
The relevant questions are what is growing, why it is growing and whether the growth strengthens the underlying business.
16. Institutional memory and the continuity of expertise
Specialist underwriting businesses may depend on the experience of a relatively small number of individuals. Their knowledge can include the reasoning behind previous decisions, the history of a market, the significance of exceptions and lessons learned from claims and broker relationships.
Some of this knowledge may be documented; some may remain tacit or difficult for others to retrieve.
The framework considers how decision histories, claims, referrals, exceptions, market observations and lessons learned might be preserved and made more accessible. Its purpose is to strengthen continuity and organisational learning, not to suggest that professional experience can be completely captured or replicated by a machine.
Institutional memory is a strategic asset only when preserved information remains intelligible, appropriately governed and useful to future decisions.
17. Governance, accountability and professional control
The use of artificial intelligence in underwriting can have material commercial consequences. Within delegated authority arrangements, the implications may extend beyond the MGA to the capacity provider and other parties to the insurance relationship.
The framework emphasises the importance of:
- Appropriate data quality and provenance.
- Testing and validation proportionate to the intended use.
- Monitoring of performance and limitations.
- Transparent reasoning and evidence where practicable.
- Auditability and clear responsibility for consequential decisions.
- Suitable referral and escalation arrangements.
- Review when assumptions, data or operating conditions change.
The level of control should reflect the nature of the application and the consequences of error. Entrepreneurial agility and rigorous governance are not inherently incompatible.
18. Human intelligence and machine intelligence
The framework does not assume that human and machine intelligence are interchangeable.
Machines can process and compare large volumes of information, retrieve records, monitor patterns and identify anomalies at a scale that would be impractical for an individual. Human professionals contribute contextual understanding, judgement under uncertainty, commercial negotiation, ethical interpretation and accountability.
Neither description implies that machines are infallible or that human judgement is free from error. The important question is how the respective strengths and limitations can be understood and combined within a responsible operating model.
BIONIC INTELLIGENCE® is concerned with engineering that relationship deliberately.
19. Technology neutrality
BIONIC INTELLIGENCE® does not prescribe artificial intelligence as the answer to every underwriting or organisational problem.
Depending on the circumstances, a better process, improved data, clearer authority, stronger training, more effective communication or a simpler operating model may offer greater value than a new technological system.
The consultancy begins with the MGA's objectives and considers the appropriate combination of people, processes, information, conventional technology and artificial intelligence.
This reflects the wider principle of X from GENERAL INTELLIGENCE PLC: artificial intelligence is a means, not an end.
20. The BIONIC INTELLIGENCE® Test
The published framework proposes that each intelligent capability should be tested against substantive questions, including:
- Does it improve the underwriter's perception of risk?
- Does it increase the quality of risk selection?
- Does it support pricing without obscuring professional judgement?
- Does it improve the understanding and oversight of delegated authority?
- Does it help reveal portfolio accumulation earlier?
- Does it improve the allocation of human attention?
- Does it strengthen the use of broker intelligence?
- Does it accelerate learning from claims?
- Does it preserve and extend institutional memory?
- Does it identify opportunities as effectively as it identifies risks?
- Does it strengthen the MGA's relationship with its capacity providers?
- Does it create an underwriting advantage that is difficult for a conventional organisation to reproduce?
The questions are intended to establish whether a proposed capability contributes to the MGA's purpose. They are not a claim that every application will pass the test or deliver a predetermined result.
21. About X from GENERAL INTELLIGENCE PLC
X is the trading name and registered trade mark of GENERAL INTELLIGENCE PLC, a Scottish company established in 1896.
The company provides artificial intelligence consultancy addressing the strategic, technical and institutional questions involved in the responsible use of intelligent systems. Its work considers not only what technology can do, but whether its use is appropriate, commercially justified, properly governed and capable of contributing lasting value.
Its approach is founded on five principles: long-term owner value, contribution, simplicity, ordinary decency and political neutrality.
The company maintains a wider portfolio of intelligence-related intellectual property and consultancy frameworks for specialist sectors.
22. Terminology
Artificial intelligence: Computational systems and methods capable of performing tasks associated with functions such as analysis, prediction, classification, language processing and pattern recognition.
Machine intelligence: The analytical and information-processing capabilities of computational systems, considered in relation to their role within a wider organisation.
Bionic intelligence: Within this framework, the deliberate integration of human and machine intelligence to extend professional and institutional capability.
Managing general agent (MGA): A specialist insurance business operating under authority delegated by an insurer or other capacity provider, with its precise functions determined by its contractual and regulatory arrangements.
Delegated authority: Authority granted to an MGA or other intermediary to perform specified insurance activities within agreed limits and conditions.
Underwriting appetite: The risks, classes of business and terms an organisation intends to accept, informed by its strategy, expertise, risk tolerance and commercial objectives.
Institutional memory: The retained and accessible knowledge of an organisation, including decision histories, experience, reasoning, exceptions and lessons learned.
23. Important clarification
BIONIC INTELLIGENCE® is a consultancy framework. This announcement describes its purpose, principles and potential applications; it does not announce a new software platform, claim that every capability described has been implemented at a client organisation, or report independently verified performance improvements.
The practical value of any proposed application will depend on the MGA's objectives, data, systems, operating model, governance and implementation. Artificial intelligence does not remove uncertainty, guarantee underwriting profitability or eliminate the need for professional judgement and accountability.
24. Closing statement
The central proposition of BIONIC INTELLIGENCE® is that an MGA's future should not be measured by the extent to which it automates its work, but by the extent to which it becomes more capable of understanding and managing the risks it chooses to insure.
Its ambition is to connect specialist judgement with broader analytical reach, individual decisions with portfolio intelligence, claims with underwriting learning and personal expertise with enduring institutional memory.
Artificial intelligence is the means. BIONIC INTELLIGENCE® is the capability. Underwriting judgement is the discipline. Sustainable underwriting advantage is the objective.