The Allianz Group is one of the world’s leading insurers and asset managers, active in almost 70 countries and serving around 97 million private and corporate customers*. Allianz customers benefit from a broad range of personal and corporate insurance services, ranging from property, life and health insurance to assistance services to credit insurance and global business insurance. Allianz is one of the world’s largest investors, managing around 764 billion euros** on behalf of its insurance customers. Furthermore, our asset managers PIMCO and Allianz Global Investors manage about 2.0 trillion euros** of third-party assets. Thanks to our systematic integration of ecological and social criteria in our business processes and investment decisions, we are among the leaders in the insurance industry in the Dow Jones Sustainability Index. In 2025, over 156,000 employees achieved total business volume of 186.9 billion euros and an operating profit of 17.4 billion euros for the Group.
For People. For Pets. For Trust: Responsible Use of AI at Allianz
Customer-first, with trust as the compass
At Allianz, the customer is at the center of everything we do. The same applies to our approach to AI: we’re using it with our customers at the heart of our strategy. We analyze, automate and personalize processes along every touchpoint with our customers. At the same time, we rigorously protect customer data, safeguard Allianz’s digital sovereignty and leverage human capabilities where they create the most value.
This customer-centric compass shapes how Allianz uses AI to drive growth and resilience. Efficiency and cost effects are important, but they are intentionally treated as outcomes of a trust-based AI strategy, not its starting point.
Governance built early
“Allianz introduced principles for responsible AI years before the EU AI Act required them”, says Philipp Raether, Chief Privacy & AI Trust Officer at Allianz. “This was a deliberate leadership decision: We wanted to scale AI without risking trust, reputation or business capability. We built governance and accountability in parallel to technological progress. It’s the right thing to do for our customers.”
Allianz established a Group Data and AI Trust Advisory Board , building on the Group Data Advisory Board founded in 2021, to advise the Board of Management of Allianz SE on data ethics and the responsible and compliant use of artificial intelligence.
Eight principles that translate trust into practice
Allianz principles for responsible AI
- Prohibited AI: Allianz does not develop or use AI systems that violate law or fundamental values, including prohibited practices under the EU AI Act, e.g. AI that manipulates individuals, exploits vulnerabilities, or otherwise poses unacceptable risks to fundamental rights.
- Transparency: Customers are informed when and for what purpose AI is used, including when they interact with AI systems.
- Accountability, accuracy and proficiency: AI might hallucinate, which is why accuracy is a responsible AI requirement: inaccurate outputs can lead to misunderstandings or sub‑optimal decisions and therefore need to be appropriately managed to ensure reliable and trustworthy use.
- Security and resilience: Technical robustness and cyber-security across the full lifecycle, including protection against misuse and failures.
- Non-discrimination: AI use cases may create a risk of discrimination when biased data leads to unfair outcomes for certain individuals or groups. This risk is mitigated by identifying, preventing, and correcting bias in the data and use of AI.
- Data privacy: AI systems process data, which is why data privacy is a responsible AI requirement: only data that may be lawfully processed under applicable data protection laws is used by or made available to AI.
- Data governance: Data governance is essential because AI relies on data, and clear rules on data quality, origin, access, and use are necessary to ensure reliable, compliant, and trustworthy AI outcomes.
- Human oversight: AI is subject to human‑in‑the‑loop oversight, meaning designated humans monitor its use and can intervene or correct outcomes that are not aligned with Responsible AI principles.
Drawing red lines: what Allianz will not do
Examples of excluded applications include:
- Social scoring of individuals based on behavior, social characteristics or inferred traits.
- Emotion recognition to assess employees at work.
- Manipulative AI that exploits vulnerabilities to influence people.
- AI designed to circumvent legal or regulatory safeguards (prohibited under EU rules and not used outside the EU either).
A streamlined lifecycle: from ideas to operations and beyond
Responsible AI in action: two examples
1) End-to-end claims processing in property & casualty (P&C)
A tangible example is automated claims processing for high-volume, low-complexity claims. Customers (or their agents) upload documents such as invoices. Optical character recognition extracts relevant information, AI models validate key fields (for example invoice date and policy number) and compare them with policy data. If the system is uncertain, cases are routed to human experts.
In Allianz pet insurance in Germany, fully automated processing accounted for 49.7% of claims in 2025, leading to a payout time of a few hours after the filing of the claim for simple, everyday claims. Transparency of the use of AI is ensured through updated privacy notices that explain when cases may be handled fully automatically, and customers always have access to a human contact to avoid being trapped in a “AI loop”.
2) Agentic automation for food spoilage claims during natural catastrophes
In Australia, Allianz uses agent-based automation to handle food spoilage claims, a frequent small claim type during natural catastrophes. Several task-focused agents work together to complete subtasks and accelerate settlement from days to hours. The approach is designed for peak loads: in 95% of cases, the maximum payout requested is AUD 500.
For claims under AUD 500, overall processing time was reduced from around seven days to less than one day. Human oversight remains decisive: potential rejections are escalated to experienced claims handlers, and employees receive targeted training to interpret and challenge AI recommendations. Management dashboards compare AI outputs with actual claim outcomes to detect deviations early.
Investing in technology and people
Beyond algorithms: trust as the true advantage
Further information
More news
About Allianz
* Customer count reflects Allianz customers in consolidated entities that are part of the customer reporting scope only.
** As of December 31, 2025.