What Happened... to all the Brokers?
CAII Memo #1729 June 14, 2032
Disclaimer: The views and opinions expressed in this blog are entirely my own and do not necessarily reflect the views of my current or any previous employer. This article is literally a work of science fiction. This blog may also contain links to other websites or resources. I am not responsible for the content on those external sites or any changes that may occur after the publication of my posts.
End Disclaimer
Disclaimer #2:
Hey- This is the first article in a pre-mortem fictional series called “What happened…?” to help me think about the possible effects and ramifications of AI in my current industry.
(e.g.- a pre-mortem imagining how AI fails in (re)insurance.)
A pre-mortem to think ahead and to help stave off or avert potentially bad outcomes.
A type of proactive mental red-teaming.
Science Fiction
Science Fiction?
Disclaimer #2 End:
(A rolling scroll like before every Star Wars movie):
It’s summer, June 2032.
The (re)insurance brokerage industry has suffered considerable losses (along with several reinsurance and insurance underwriting companies).
Several brokerages have gone out of business and the remaining players have consolidated in a bid to retain power against the rising tide of AI disintermediation in their industry.
Technological advancements have affected yet another “human touch” industry.
What follows is a brief outline of the events that contributed to the disintermediation.
(End Star Wars-like scroll)
INT. LOUIS’ APARTMENT - NIGHT
We open on Louis Friend, mid twenties, nerdy, junior AI regulator for the government created Committee for AI in Insurance (CAII).
He is alone at his laptop, sitting at his desk in his spartan city apartment. A desk lamp lights the area in front of him. A cat works to get comfortable on a sofa in the background. He is staring at his computer screen.
He lifts his glasses and rubs his brow. It’s been a long night preparing for tomorrow’s presentation before the Committee.
We zoom in, over his shoulder on his computer screen as he reviews his prepared statements for tomorrow’s meeting, and the rest dissolves…
Date: June 14, 2032
By: Louis Friend
Memo: A Timeline and description of the Recent Events Affecting the (Re)insurance Brokerage Industry
Introduction
The insurance and reinsurance industries, once dominated by human brokers and agents, have witnessed a seismic shift in the last decade.
The year 2032 marks a pivotal point: the shuttering of 2 large multinational brokerages, and the merging of the remaining 3 large players into “The BrokerAIge™”, a largely AI-driven dark pool that matches risks between customers, and uses Quantum Super Language Models (QSLMs), aka “Supers”, to automate the process from start to finish.
This transformation may be attributed to the rapid advancements in Artificial Intelligence (AI), machine learning, and generative AI technologies and in the subsequent lack of adoption by the industry until it was too late.
Committee members, please allow me to review the key events and factors that led to this dramatic change.
June 12, 2017: Attention is All you Need Paper Published, introducing Transformer neural net architecture.
September 19, 2019: OpenAI publishes research on GPT-2 language model.
November 30, 2022: OpenAI introduces ChatGPT using GPT-3.5.
2024: Early Industry Adoption and Skepticism
A very small subset of (re)insurance companies and brokerages begin the adoption and implementation of generative AI tools into their workflows.
In light of the fact that few companies have dedicated Applied Data Science and AI (ADSAI) teams, several companies choose to buy rather than build out internally. (One popular 3rd party vendor is brought in by a number of underwriting companies- a fact that will contribute to the collective D&O Winter of 2029 that occurs when those same companies find that they share the same algorithmically selected, pooled bets.)
These initial years are marked by a cautious adoption. Insurance and reinsurance companies begin experimenting with AI for data analysis and risk assessment. Despite skepticism, AI tools show potential in simplifying complex underwriting processes and predicting risks with unprecedented accuracy.
2026: Continued Automation and Chief Algorithmic Underwriting Officers
Continued automation occurs throughout the industry which in turn allows the newly cleaned and standardized data to be fed into the the current crop of multimodal llm models of late 2025 including GPT 5- CD (a nod to Star Trek’s Commander Data) and EchoGen Lab’s new Oppy 1 model.
Better, cleaner data, and a newly agreed upon universal insurance data schema, allow for insurance and reinsurance companies to talk directly with one another with less need for a data transport intermediary.
Several Chief Algorithmic Underwriting Officers are hired industry wide by insurance and reinsurance companies, a role that arguably in 2026, is instituted 2 years too late.
2028: EchoGen’s new Oppy model achieves Full Agency status.
(“Full Agency” status having replaced the now long defunct Turing Test.)
An offshoot of EchoGen’s IndustryTuned™ models, called Leadenhall 12™ (after original Lloyds of London location), in a first, completely recreates a reinsurance company portfolio from scratch- all lines of business-having trained on a combination of generative AI created synthetic data and actual company audio and video recordings, garnered from 18 months of office, meeting, and VOIP conversations.
2029: The Regulatory Shift
In a move likened to the individual mandate of the Affordable Care Act of March 2012, the Supreme Court Orders that, for regulatory, safety, and accuracy purposes, several underwriting lines must be handled both completely algorithmically, and directly between buyer and seller.
( Note: a similar mandate for autonomous driving, to increase driver safety, will occur in March of 2031)
The government introduces the first of many line-specific “risk matching pools” similar to dark pools that started in the 1980’s on Wall Street and grew further after the introduction of the SEC’s Reg NMS in 2005.
Insurance and reinsurance companies begin to grow their technical infrastructure to tailor to these direct, brokerless systems. Those lines not directly mandated by the government increasingly switch to this type of direct, algorithmic risk matching.
Several (re)insurance lines now run in an automated fashion.
Broker commissions collapse.
2029: The D&O Winter and CAT resurgence
Several insurance companies exit the D&O business entirely due to sustained losses (see 3rd party vendor pooled bets). This marks a formal capitulation in the line for traditional insurance and reinsurance companies, who have for years been losing market share to quantitative hedge funds and other 3rd party capital employing the most state of the art (SOTA) AI tools to underwrite algorithmically.
A line many thought would experience continued profound difficulties, property catastrophe underwriting (CAT), experiences a newfound resurgence given advances in artificial intelligence weather models which allow for both extremely granular real-time and seasonal predictions. These advances, coupled with increasing CAT activity globally due to “extreme weather events”, have created robust two-sided trading in several new standardized and over-the-counter (OTC) trading markets.
Some examples:
Micro-climate insurance products based on hyper-localized data.
Seasonal Shift insurance- Coverage for businesses affected by shifting seasonal patterns due to climate change, such as altered tourist seasons or agricultural cycles.
Real-time event AI-Optimized Reinsurance Layering
Dynamically adjusting, algorithmically driven barrier knock-in and knock-out options for live CAT events
Given their increased frequency, certain states allow individuals to bet on various parametric aspects of weather and CAT events. (Similar to 2018, when the Supreme Court struck down the Professional and Amateur Sports Protection Act (PASPA), allowing each state to create and enforce its own sports betting laws.) The increased tax dollars help the coastal (and increasingly inland) states offset losses from weather events and decreased tourism revenue.
2030: The global “MataLeão” cyber attack occurs, beginning February 26th.
(Please see details in Appendix 3 and CAII memo #1497)
Cyber underwriting in its current form is no longer tenable.
2032: Closures and Consolidation
By 2032, AI and machine learning had advanced to a point where human intervention becomes largely unnecessary in the (re)insurance brokerage process.
2 large multinational brokerage firms close their doors, and the remaining 3 largest players merge to form their own AI driven dark pool called “The BrokerAIge™”.
New lines of business such as Social Credit insurance, real-time CATalyst™ coverage, and AI Personal Companion Liability insurance help to soften the blow, but the contraction is irreparable.
Conclusion and Impact on the Workforce
This paradigm shift led to a massive displacement of jobs. However, it also opened new opportunities in AI development, data analysis, and cybersecurity within the insurance and reinsurance sectors.
Those companies that began to adopt AI early and reskill their employees have been able to find new ways to make money. Those that did not suffered fates similar to what befell pockets of the banking and financial services sectors around 2030.
Looking Ahead
The disintermediation of (re)insurance brokers is a testament to the transformative power of AI. It raises important questions about the future of employment and the need for upskilling in an increasingly automated world.
Esteemed members of the Committee, as we embrace this new era, the key challenge remains: balancing technological advancement with its societal impact.
This concludes my prepared statement. I thank the Committee for their time and look forward to answering any questions they may have. As per protocol, I will be switching my AI assistant off Corporeal Mode™ during the question period.
INT. LOUIS’ APARTMENT - NIGHT
We slowly zoom out from Louis’ screen and the blinking dot of his cursor.
Louis closes his laptop and exhales. Time for bed. The cat leaps off the sofa cushions and follows Louis down the hall past a framed poster he has hanging on the wall outside the entrance to his room. We see the light turn off in Louis’ room and the camera begins to slowly zoom in on the poster on the wall, until we are right in front of it
There, in bright red letters the poster reads:
Don’t Slow Down
FADE OUT.
CREDITS.
END.