Industries / Finance and business services / Insurance / Insurance brokerage
Placing policies
Placing is getting the chosen insurers to commit to a risk on agreed terms, and recording the contract.
7
Milestones achieved, from 1810 to 2025.
3
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1810
Followers trust a leader
In a subscription market, one large risk is shared among many underwriters: a lead underwriter sets the terms and others follow by taking shares. In evidence to a House of Commons committee in 1810, the broker James Barnes said that underwriters "who see a policy begun by a person in whose judgement they have confidence, will follow his example", as quoted, in paraphrase, by the Lloyd's Market Association (LMA) in 2026. Brokers still go "to see underwriters at their 'boxes'" at Lloyd's, in the words of a 2010 Gresham College lecture.
Credit
James Barnes; the House of Commons select committee
2007
Terms agreed before cover
In December 2004 John Tiner, chief executive of the UK Financial Services Authority (FSA), named the problem "deal now, detail later": cover often began before the contract wording was agreed. He gave the market two years to fix it or face regulation. In January 2007 the market told the FSA that 90% of contracts in the subscription market and 88% of the rest were achieving "contract certainty", and from 1 November 2007 the Market Reform Contract (MRC), a standard slip, "became a mandatory standard for the placement of general insurance contracts". Lloyd's still requires it.
LessonA deadline backed by the threat of rules got the market to fix the problem itself, "without a single new rule being introduced", in Tiner's words.
Credit
John Tiner, FSA; the Market Reform Group, now the London Market Group
2016
A fair presentation
The UK Insurance Act 2015 replaced the old duty to disclose with a duty of fair presentation, for contracts entered into or varied after 12 August 2016. A business must disclose "every material circumstance which the insured knows or ought to know", in a way "reasonably clear and accessible to a prudent insurer". The official notes warn against "data dumps", "an overwhelming amount of undigested information", and against "an overly brief or cryptic presentation". Brokers, who usually prepare the presentation, no longer carry a separate duty of their own.
Credit
UK Parliament
2019
Facilities under review
A broker facility is a standing agreement in which insurers take an agreed share of business the broker brings, often without pricing each risk themselves. In 2013 Lloyd's warned syndicates that such deals involve "delegation of underwriting authority to the broker". In February 2019 the UK Financial Conduct Authority (FCA), in a study of 73 wholesale brokers and 49 underwriters, said it had "not found evidence of significant levels of harm to competition that merit the introduction of intrusive remedies", but found that broker pay "for like-for-like policies" was "between 4-6% higher via facilities". Aon says its Client Treaty facility will fill 35% of each eligible placement in advance from 1 January 2027, up from 28.5%.
LessonPre-arranged capacity speeds placing but gives the broker a reason to send business to its own facility, so regulators watch the pay.
businessinsurance.comfca.org.ukfca.org.ukreinsurancene.wsreinsurancene.ws
Credit
FCA; Lloyd's
2020
Placing leaves the room
PPL, the London market's electronic placing platform, began trading on 11 July 2016. In the first quarter of 2019 syndicates accepted 45% of in-scope risks through it, against a target of 40%. Lloyd's closed its underwriting room "with effect from 16:00 on Thursday, March 19, 2020", after testing whether "the market could trade electronically via PPL". Lloyd's later ended its mandate for electronic placing, saying the share was "now over 90% for in scope business".
lloyds.comintelligentinsurer.comcarriermanagement.comlloyds.com
Credit
the London Market Group, which delivered PPL; Lloyd's
2021
An algorithm follows
Ki, a Lloyd's syndicate set up by the insurer Brit, is "a follow-only digital business, built by Google Cloud and accessed directly by brokers", developed with University College London and backed by Blackstone and Fairfax. It does not lead risks; it follows the terms set by other underwriters, and its algorithm was built to "evaluate Lloyd's policies and automatically quote for business". Its first risks incepted in January 2021. Its gross managed premium was $1.11 billion in 2025, by its own results. Ki calls itself "Lloyd's first fully digital and algorithmically powered syndicate".
Credit
Brit; Google Cloud; University College London
2025
German court backs the lead
In German co-insurance, several insurers share one policy and a lead insurer handles it for them. In a judgment of 7 January 2025 (ref. 9 U 37/20), the Hanseatic Higher Regional Court held that the co-insurers had "relinquished their powers of co-determination and control", so the lead's settlement bound them, by the law firm Heuking's account.
Credit
Hanseatic Higher Regional Court
Watch for
Placing in days, not weeks
Marsh says its Broker WorkBench, an AI system for London specialty risks, will "reduce placement times from the industry average of two-to-four weeks to days or hours". It lets brokers "bind contracts for lead markets, digital followers and pre-arranged capacity", and Marsh's brokers still approve each placement.
Achieved wheninsurers, clients or an auditor publish measured placement times or volumes.
Claimed by
Marsh, September 2026
Digital followers multiply
Ki added Tokio Marine Kiln as its fifth capacity partner, bringing "over $200 million of partner capacity", and AEGIS London and Augentic launched a digital follow consortium. Insurance Business wrote that after Lloyd's central programme fell short, "the market cannot wait for a centralised solution to arrive".
Achieved whenLloyd's or an independent analyst publishes the share of following lines written digitally.
Claimed by
Ki, April 2026; AEGIS London and Augentic, 2026
One data record per contract
The Core Data Record (CDR) is the London market's standard set of data fields for a contract. The London Market Group's Data Council extended it to the first notice of a claim and to delegated authority, and said "Work is already underway to incorporate the expanded CDR into the next version of the Market Reform Contract." No figures on its use were found.
Achieved whenLloyd's or the market publishes the share of risks placed with a complete CDR.
Claimed by
London Market Group Data Council, October 2026
Open
01
Slips turned into data, tested
A public test set of 300 made-up MRC slips in mixed formats (Word files, PDFs, scans, slips with endorsements), each with a hand-keyed answer key of CDR fields, and an open-source converter that turns a filled slip into those fields. A public script scores any converter on the set.
The briefCredit · unclaimed02
A fair presentation checker
100 made-up companies, each with a full fact file and a broker-style presentation of its risk to insurers. The presentations contain 400 planted omissions of material facts, such as an earlier fire or a sprinkler system out of service, and 100 material facts buried in long attachments. A tool reads both and lists each fact that the presentation fails to make "reasonably clear and accessible".
The briefCredit · unclaimed
Open 01 · the brief
Slips turned into data, tested
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public test set of 300 made-up MRC slips in mixed formats (Word files, PDFs, scans, slips with endorsements), each with a hand-keyed answer key of CDR fields, and an open-source converter that turns a filled slip into those fields. A public script scores any converter on the set.
Why it is still open
V7, which sells slip extraction, says its extracted values are "mapped to the Common Data Record schema" but publishes no accuracy figures, and tells buyers to test on their own slips because "It is the only benchmark that transfers." No open-source MRC parser and no public slip test set were found.
Built with what exists
the CDR field list and MRC template published by the London Market Group, subject to their terms of use; document-reading AI models such as Claude and Gemini; Tesseract or docTR for text in scans; JSON Schema for the answer key.
First step
Get the current CDR field list and MRC template, check what their terms allow, and draft 20 slips by hand.
Done looks like
At least 95% exact matches on the mandatory CDR fields, and at least 95% of deliberately missing fields flagged as missing rather than filled in. The test set, answer key and scorer are public, with 100 slips held back.
Open 02 · the brief
A fair presentation checker
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
100 made-up companies, each with a full fact file and a broker-style presentation of its risk to insurers. The presentations contain 400 planted omissions of material facts, such as an earlier fire or a sprinkler system out of service, and 100 material facts buried in long attachments. A tool reads both and lists each fact that the presentation fails to make "reasonably clear and accessible".
Why it is still open
The law firm Browne Jacobson wrote in July 2026 that "the ability to identify material disclosures buried within a lengthy submission is critical", and that "the courts have not yet had occasion to specifically address" the effect of AI. No public test set of risk presentations was found; brokers' associations publish checklists for people.
Built with what exists
long-context AI models such as Claude and Gemini; section 3 of the Insurance Act 2015 and its explanatory notes on "data dumps" as the marking scheme (see the Note "Seven more flagged"). All companies are made up, and the tool gives no advice to real clients.
First step
Write ten fact files and presentations with 40 planted omissions, and have two people mark them.
Done looks like
The tool finds at least 90% of the planted omissions, and no more than 10% of its flags are false, as judged by two reviewers working separately. The data and marking scheme are public.