An Allstate advert shows a 'bear' raiding autos for the coolbox contents left inside and warns consumers to make sure their policies cover the risks they are likely to be exposed to avoid potential mayhem. Now it is suffering mayhem itself.

Allstate reported a 2Q loss of $4.42 per diluted share earlier this week, with its stock down around 3 percent since last Friday. (The stock is off nearly 20 percent year to date.) Progressive’s stock has fared better after it eked out 57c of 2Q per-share profit, but that was because it cut spending on advertising; it still lost money on its underwriting businesses.

Repair costs, litigation, and accident severity have all added to the woes of auto insurers whilst 

Weather blowing in severe thunderstorms and tornadoes caused lots of damage and multiple deaths this spring; the US had $40bn of insured catastrophe losses in the first half of this year, the third-highest on record after 2011 and 2021, according to Aon, an Illinois consulting firm and reinsurance broker:


We have seen similar issues in the UK with high-profile carriers suffering losses and even CEOs being shown the door.

Premiums are rising but still lagging behind claims inflation. When they catch up with current trends will other shocks come to hit bottom-line profits? Underwriters seeking more data on battery electric vehicles (BEV) as sales increase and battery replacement/repair costs result in total loss decisions even after just a year, climate change accelerating and outliers becoming the norm all make predictions harder.  

Some may hope for LLMs and generative AI to provide the answer but remember that these tend to be backward-looking, under-value outliers and need considerable oversight and governance. They favour the view of the average practitioner rather than the intelligence of outstanding practitioners. They carry forward bias. Cigna has already been taken to court over alleged bias resulting from reliance on AI without the necessary oversights.

Data, data, and world-class data management is key to managing the present and anticipating the future.  Was ever thus.

Visit Instech's website for podcasts and face-to-face meetings where insurers and insurtechs meet to discuss how to source data for today's challenges. Even with the right data, set clear goals for leveraging this data.  If insurers have not successfully deployed RPA, machine learning, and narrow AI use cases they cannot expect to leverage genAI.

GenAI is the darling of the moment but take a moment to consider the thoughts of David M Brear CEO of 11:FS

"I'm sitting in the 11:FS headquarters in Broadgate, London. What's going on around here is a perfect metaphor for the industry.

They have taken down two massive buildings and started to rebuild them. They took them right down to the ground, and we witnessed a flurry of activity, diggers, cranes, etc. Every little kid's fantasy playground of big machines!

That's not quite right actually; they haven't just taken it down to the ground; they have taken it down far below the ground level! For months now, they have been reconstructing the foundations, the things that hold up the building that will be built upon the ground.

They know that if they want to build a tall tower, they need a foundation that gives them stability. They knew the building they wanted wouldn't happen without those foundations being right.

This is the issue with large fi's (and insurers) ... they keep chasing the DLT, GenAI, Metapherse buzzword bs things instead of truly getting their foundations right and their operating model machine optimised for cost-effective and fast change.

Stop putting your icing on the plate people. Make the cake good first!"

But with the mayhem impacting insurers today they had better follow best-in-class and start on the path of leveraging all the data management and AI tools available to face the challenges described above. That's after making the cake good first of course as without that sound foundation how can you decide the goals you want to achieve, the data sets needed and training plans, the resourcing, and controls to leverage GenAI and not suffer the downsides? 

Further Reading

Impact of BEV Adoption on the Repair and Insurance Sectors

Motor insurers endure worst underwriting conditions in a decade

Mitigating Home Claims Inflation

The state of AI in 2023: Generative AI’s breakout year