Artificial intelligence has become the explanation of choice for workforce cuts in insurance this year. In May, Acrisure said it would eliminate 2,250 jobs, roughly 11% of its global workforce, pointing to advances in AI and automation, its second AI-linked reduction in seven months. Allianz Partners is reportedly preparing to cut 1,500 to 1,800 travel insurance roles over the next 12 to 18 months as AI reshapes its operations.
A new survey suggests the people making these decisions may have a rosier view of AI’s progress than the senior managers just below them.
BlackRock’s 2026 Global Insurance Report which has surveyed 463 senior executives in 31 countries in July and August, together overseeing more than $24 trillion in insurance assets. Overall, 60% said their firm has at least one AI use case running at scale or embedded in day-to-day operations. That figure hides a split. Among C-suite respondents it rose to 73%. Among managing directors, directors and vice presidents, it was 51%.
The survey doesn’t explain the gap, and it is a comparison across respondents, not between bosses and their own teams. One possible reading is that executives have a wider view of what’s happening across the business. Another is that a project counted as “scaled” in the boardroom still looks like a pilot to the people running it. In either case, the gap matters at a time when headcount decisions are being justified by how far AI has come.
The survey shows AI furthest along in core insurance work rather than investment. Some 54% of respondents said AI is scaled or operational in underwriting and pricing, 53% in claims, fraud and servicing, and 52% in compliance and reporting. Those are the back-office and processing areas where job cuts tend to fall.
Investment functions are further behind. Only 38% have scaled AI in investment research, with the remaining 62% still piloting or exploring. In security selection and credit monitoring, the scaled figure drops to a third. Based on the report’s charts, the gap between C-suite and other respondents appears widest in underwriting, claims and compliance.
Some carriers are reporting measurable gains. Aileen Mathieson, interim CIO of the UK’s Pension Insurance Corporation, said in the report that using AI for deep credit analysis had already produced a 20% efficiency improvement in that area this year.
Read next: AI is cutting insurance jobs. The industry is just starting to say so
Spending is still rising. AI is the top technology investment priority for the second year in a row, and 62% of insurers expect their investment teams to spend more on it over the next 12 months. That fits with Accenture research showing insurers pushing AI past the pilot stage despite skills shortages, which found 34% of insurance organizations deploying AI agents across multiple functions.
The obstacles to scaling are mostly about control, not cost. In BlackRock’s survey, regulatory and compliance concerns led at 49%. Cybersecurity and operational risk followed at 46%, then data quality at 45% and integration with existing systems at 40%. Cost came in at 24%, and only 5% cited a lack of senior management or board support. In other words, leadership enthusiasm isn’t the constraint.
Company size changes the problem. Insurers managing $1 billion to $10 billion were almost twice as likely as those above $100 billion to cite systems integration (46% versus 24%). The largest firms were more worried about cybersecurity (57% versus 38%).
Regulators are watching too. The NAIC has set up a working group on the ethical use of AI that is examining how insurers use algorithms in pricing and claims.
Jeff Mauro, CIO of Fortitude Re, said in the report that for most firms the barrier “is not the models themselves; it is data quality, governance, and controls.” He warned that insurers that skip those basics risk “scaling errors rather than insight.”
Read next: Insurers push ahead with AI despite skills gap
The report’s own commentary pushes back on the idea that AI replaces people, at least in investment work. BlackRock’s Aladdin team argues that in a regulated business like insurance, AI should earn autonomy step by step, with a named person accountable at each decision point. Martin Zingg, group CIO of Swiss Re, said the company’s ambition is “not to replace human judgment but to strengthen it.”
That is a different message from the cost-cutting announcements elsewhere in the industry. Whether it holds as adoption speeds up is one of the bigger open questions for 2027.
Insurers aren’t only using AI. They are increasingly being asked to pay for it. A paper for the Brookings Papers on Economic Activity projects $10.3 trillion of AI infrastructure investment between 2025 and 2032, and warns that the risks are moving toward less transparent financing such as private credit and securitization. BlackRock’s infrastructure credit team estimates 70% to 80% of AI-related capital spending will be debt-financed. It argues that insurers’ long-dated liabilities make them a natural lender, while also cautioning that carriers may already hold the same hyperscaler exposure through their public bond portfolios.
Read next: NAIC targets $1.2 trillion in insurer private credit with tighter solvency rules
Beyond AI, the report shows an industry holding its overall risk steady while reshuffling what sits inside its portfolios. Some 82% of insurers plan to keep investment risk unchanged. North American confidence in a soft landing fell from 83% to 63% over the year, while inflation remained the top macro concern, cited by 62%.
Appetite for private markets has cooled but remains strong. Some 76% plan to increase allocations, down from 94% last year, and private debt leads the list at 41%. The top worry about private debt is regulatory and capital treatment, cited by 56%, as privately placed bonds approach half of US life insurers’ bond holdings. And 84% now run a hybrid model that combines in-house and outside asset management.
BlackRock titled this year’s report “Different by Design,” and its central argument is that insurers are taking very different approaches to the same pressures. The AI findings suggest the differences don’t stop at the company level. Even within insurers’ senior ranks, executives disagree about how much has actually changed.
