Allianz Group is one of the world’s leading insurance and wealth management companies, serving approximately 125 million individual and business customers in approximately 70 countries.
According to an Allianz media release, the company reported total business revenue of $208 billion and operating profit of $18.5 billion in 2024.
Although Allianz does not disclose the exact amount of its investment in artificial intelligence, its commitment is evidenced by extensive infrastructure development, workforce expansion, and groundbreaking implementation of sophisticated AI systems.
The company operates AllianzGPT, an internally hosted generative AI platform launched in September 2023, serving more than 60,000 employees as of early 2025, with deployment coverage across all 158,000 employees worldwide.
Evident AI’s June 2025 Insurance Index study found that Allianz employs approximately 10% of the AI workforce of 30 major North American and European insurance companies, underscoring the concentration of talent and organizational commitment to AI development.
This analysis focuses on two AI use cases that directly support Allianz’s operational and strategic objectives.
- Automating claims processing for low complexity events: Deploying agent AI improves customer satisfaction by reducing claims processing time while maintaining human oversight of payment decisions, freeing up specialized claims staff to handle complex cases.
- Accelerate fraud management: Deploy supervised learning algorithms trained on historical claims data to instantly flag anomalous claims for expert review and improve fraud detection.
Automating claims processing for low complexity events
Allianz’s traditional claims processing model operated under systemic constraints during natural disaster (NatCat) events, such as severe weather disruptions, resulting in a surge in claims.
These catastrophic events present an operational paradox for insurance companies. Customers submit high-priority claims that require complex judgments (structural damage, business interruption), while at the same time generating a large number of low-complexity claims that consume a disproportionate amount of staff attention.
In an Allianz newsletter, Thomas Bach, managing director of core insurance platforms at Allianz Technology, characterized operational constraints as “taking more than four days to process” as the claims team’s focus was on more complex claims that occurred during the NatCat event.
This operational challenge reflects broader industry issues. According to McKinsey analysis, insurers face significant workflow bottlenecks, traditional process limitations, and customer experience gaps with traditional claims processes across the claims lifecycle. These pressures and more are leading many to pursue AI-powered solutions to improve operational efficiency.
The global insurance industry processes hundreds of millions of claims each year, and NatCat events have caused extreme claims spikes.
According to Allianz internal research and cited by the Insurance Information Institute (III), 24% of U.S. commercial insurance claims were due to natural disasters. According to the III blog linked above, from 2017 to 2021, the largest loss drivers cost approximately $90 billion, and insurance companies pay out more than $48 million every day to cover these losses.
In July 2025, Allianz launched Project Nemo as a departure from traditional billing automation. Rather than strictly using a deterministic, rule-based system that follows pre-determined decision trees, Nemo introduces agent AI. Agent AI is a specialized, task-oriented digital agent that independently plans, decides, and collaborates to complete multi-step workflows with minimal human intervention up to critical decision points.
For claims related to food spoilage, for example, the system consists of seven specialized agents, according to the Allianz newsletter.
- Planner Agent coordinates the entire workflow and maintains process status throughout the claims evaluation.
- Cyber Agent enforces data security protocols and policy guardrails to prevent unauthorized access during sensitive claims processing.
- The coverage agent will check whether the claimant’s insurance policy includes coverage for food spoilage due to inclement weather (basic coverage check).
- Weather Agent integrates with weather databases to cross-reference recorded conditions with customer-described weather events to prevent fraudulent claims related to non-existent weather events.
- Fraud agents apply machine learning pattern recognition to identify statistical anomalies that suggest intentional misrepresentation and flag high-risk claims for human review.
- The paying agent will calculate the settlement amount based on your claim details, insurance terms, limits, and deductibles.
- Finally, Audit Agent generates a comprehensive overview of all agent decisions and reasoning, creating a complete audit trail for compliance, quality control, and human review.
The entire technical process is completed within five minutes, after which a human claims expert reviews the audit summary and makes a final payment approval decision.
This architecture embodies what Maria Jansen, Chief Transformation Officer at Allianz Services, describes as Allianz’s core human-based AI governance principles. “With Project Nemo, the AI agent supports the team by making recommendations, but the ultimate responsibility always lies with the claims professional.”
Project Nemo launched in Australia in July 2025 and achieved full operational deployment within 100 days.
According to Allianz Newsletter and Australian Insurance Coverage, Project Nemo has produced the following business results:
- 80% reduction in claim processing and settlement time: For eligible food spoilage claims less than $327, processing time has been reduced from days to one day to hours.
- Near-instant technical processing: A complete seven-agent workflow runs from claim submission to readiness for human review in less than five minutes.
- Scalability in times of crisis: By automating low-complexity claims, Nemo frees up human staff to focus on complex cases, dramatically increasing response capacity during NatCat events when claims volume spikes and time pressure increases.
Project Nemo differs from point solutions because of its modular architecture and explicit design for cross-product scalability. Previously cited press materials state that Allianz is considering deploying the agent framework to other low-complexity, high-frequency use cases, such as travel delay claims, simple auto insurance claims, and property damage assessments.
Accelerate fraudulent billing management
By 2023, Allianz’s fraud investigation team was overwhelmed by a growing number of claims and faced increasingly sophisticated fraud techniques that traditional rules-based systems could not detect.
According to the Allianz newsletter, the quantitative reality reflects this deterioration.
- In 2023, Allianz Commercial discovered billing fraud of £77.4m, up from £70.7m in 2022.
- Fraud schemes have evolved through sophistication and coordination. A proliferation of false Confirmed Claims Experience (CCE) documents (commercially equivalent to no-claims bonuses) hinted at possible identity theft for genuine businesses.
- Automotive fraud is on the decline, with “crash for cash” referrals up 25%.
The operational bottleneck was not an absolute lack of detection capacity, but rather a lack of investigative resources. Fraud experts were faced with a growing queue of suspicious claims that exceeded their investigative bandwidth, causing delays in processing both legitimate and fraudulent claims.
The crisis is widespread across the industry, with the Association of British Insurers (ABI) reporting that insurers discovered more than 98,400 fraud-related claims in 2024, up 12% from 88,100 in 2023. According to Carpe Data’s 2025 Fraud Report, traditional fraud detection methods only analyze 5% of unresolved personal injury claims, so the fundamental problem is the ability to detect fraud, not the methods used to identify it.
In response, Allianz introduced Incognito in 2023. It is a supervised machine learning-based tool designed to identify potentially fraudulent claims.
Allianz has not disclosed the system’s architecture or workflow. However, industry standard approaches are extensively documented.
Insurance fraud detection systems typically employ supervised machine learning, where algorithms are trained on past insurance claims that have been previously labeled as fraudulent or legitimate. International journal of advanced research in science, communication and technology.
According to that study and another study in the Journal of Insurance and Risk, the industry detection system workflow works as follows.
- Billing data is preprocessed. The supervised model outputs the probability of fraud, and the unsupervised model outputs the anomaly score for each claim.
- Claims are ranked by these scores and the top subset is sent to human investigators.
- Investigators label suspicious cases. These labels serve as criteria for evaluating model performance.
Importantly, Incognito identifies potentially fraudulent claims and refers them to fraud experts for thorough review and investigation, according to Allianz’s newsletter. The system does not autonomously determine wrongdoing. Instead, it acts as an intelligent triage mechanism.
According to Allianz sources cited below, Incognito has delivered the following results for the company:
- According to another Allianz newsletter, Allianz UK was able to reduce fraud by £37.7m in the first half of 2024.
- According to the Allianz newsletter, overall insurance claims fraud detections increased by 10% year over year in the first year.
- According to the Allianz report, application fraud reductions increased by 150% compared to year-to-date expectations.
