Humanyx
    AI comparison · Oct. 2026

    Which AI may insurance brokers use with client data?

    Three questions, one clear answer: which cloud AI is permitted for your industry, where your data is processed and how strong the models really are on that route.

    Humanyx guides financial service providers as a partner into an AI-enabled operating model. This overview shows all common routes, including those we do not implement ourselves.

    Your industry

    Industry rules and supervision feed directly into the traffic light.

    IndustryInsurance brokers and intermediaries

    Which data should go into the AI?

    Health data in life, daily sickness allowance and occupational pension files has its own traffic light.

    Where may your data be processed?

    Your internal policy, not the law. We show both.

    For Insurance brokers and intermediaries

    FINMA registration and professional duties, contractual confidentiality and the DPA. Health data in life, daily sickness allowance and occupational pension files has its own traffic light.

    Your result · Insurance brokers · Switzerland only
    3of 13 options fit your requirement
    • 3Green
    • 1to check
    • 9leave Switzerland
    • 0Red
    What does the traffic light mean?
    Your best model on this route
    42points, with preview 45

    Top score today: 58 with Claude Opus 5.5, available with EU processing (traffic light yellow)

    PreviewIn the preview at Phoeniqs, DeepSeek V4.1 Flash (45), GLM 5.3 (45), Kimi K3 (44) are running.

    The number is the Artificial Analysis Intelligence Index: an independent benchmark with ten demanding task sets covering agentic tasks, office work, knowledge, coding and science. Scale 0 to 100, the best model scores 58 today. As of 23 October 2026, version 4.3.2. "ca." = estimated.

    The best solutions: three proven setups for insurance brokers and intermediaries

    Hardly any organisation needs just one tool. What has proven itself is the combination of an AI employee for daily work and a secure interface for automations.

    For brokers we recommend the strict setup: a Swiss platform as the working surface and Azure Switzerland Standard (GPT-4o) for automations. Provisioned with GPT-5.1 is too expensive for small and mid-sized brokers and retires in May 2027.

    Recommended for insurance brokers

    Strict: Swiss platform plus Azure Switzerland

    • A Swiss AI platform as the working surface for all employees
    • Azure OpenAI Switzerland North (GPT-4o) or Provisioned (GPT-5.1, from approx. USD 13,000 per month) for automations
    • External models in the platform selectively disabled per use case
    Proven for everyday office work

    Pragmatic: Copilot plus Azure Switzerland

    • M365 Copilot for emails, documents and meetings, with storage in Switzerland and Flex Routing disabled
    • Azure OpenAI Switzerland North for automations with sensitive client data, such as document review or reports: for new customers, GPT-4o (until April 2027) is effectively the only option there
    • A clear policy: what may go into Copilot, what only goes through the Swiss pipeline
    When quality matters above all

    High performance: reference architecture with frontier models in the EU

    • Frontier model for quotes, comparisons and correspondence
    • Health dossiers stay on the Swiss pipeline
    • Risk assessment, contract and client information with extra care
    How this split works for brokers
    Sovereignty map

    The closer your data stays, the smaller the choice of models.

    Each dot is an option, placed by data location and by the performance of the best model on that route. The colour shows the traffic light for your selection above.

    Intelligence Index of the best model on this route (Artificial Analysis, scale 0 to 100)
    • Green
    • Yellow
    • Red
    • Preview, no SLA
    Preview at PhoeniqsDeepSeek V4.1 Flash, GLM 5.3: 45Kimi K3: 44
    16 pointsbelow the top score
    Switzerland onlyProcessing and storage in Switzerland
    No measured valueM365 Copilot: GPT models, chosen by Microsoft.
    Excluded by your requirement
    EUProcessing in the EU
    No measured valueCopilot flex routing: GPT models, chosen by Microsoft.
    Excluded by your requirement
    Worldwide incl. USANo data location guaranteed
    What does the number mean?
    As of 23 October 2026

    Artificial Analysis Intelligence Index. Independent benchmark by Artificial Analysis. All models solve the same ten demanding task sets: agentic tasks and real office work (30%), general knowledge, long documents and hallucinations (30%), coding (20%) and scientific reasoning (20%). Scale 0 to 100, higher is better. The tests are deliberately very hard: the best model scores 58 today.

    Version 4.3.2, values as of 23 October 2026. "ca." = estimated by Artificial Analysis, the independent measurement is still pending.

    Switzerland only

    In Switzerland, GLM 5.3 Flash is stronger than any available GPT model.

    At Phoeniqs with 42. Azure Switzerland only offers older GPT models: GPT-5.1 (ca. 25, Provisioned only) and GPT-4o (ca. 8). The newest GPT models only run with EU processing.

    Preview

    Kimi K3 runs in Switzerland, but only as a test.

    Phoeniqs is testing Kimi K3 (44) until 16 November 2026 without an SLA, and production use is not planned according to the documentation. DeepSeek V4.1 Flash and GLM 5.3 (45 each) are meant to go into production.

    EU

    The top score is also available with an EU guarantee.

    Claude Opus 5.5 via AWS Bedrock or Google Vertex AI with EU processing reaches 58, the highest value in the index. Traffic light Yellow: feasible with a risk assessment and pseudonymisation.

    All AI options at a glance

    The traffic light applies to your selection above. Filter by colour and open an option for the data path, models and checklist.

    How to read the traffic light

    Tailored to insurance brokers and intermediaries: FINMA registration and professional duties, contractual confidentiality and the DPA, health data in life, daily sickness allowance and BVG dossiers with its own rating. Stricter rules apply to insurers (Circular 2018/3 directly, outsourcing in the business plan): talk to us.

    • Green: proven and recommended

      Works with client data: the setup is proven, contract and governance are in place. This is exactly how you should do it.

    • Yellow: unclear, depends on the setup

      Depends on the contract, configuration and industry: feasible with a documented risk assessment, not without it.

    • Red: not possible

      Not defensible for client data in our risk assessment: no contract, no control or no documented safeguards. Often still the best choice for research, marketing and anonymised content.

    In this view the first rating applies to property and commercial client data, the second to dossiers containing health data. Health data is especially sensitive (Art. 5(c) DPA): large-scale processing requires a data protection impact assessment (Art. 22(2)(a) DPA), consent must be explicit (Art. 6(7)(a) DPA), and medical reports and claims narratives can hardly be pseudonymised reliably because diagnosis, accident date and location together identify a person. A small or mid-sized broker rarely manages this, so the Swiss pipeline is the more honest advice.

    Note on the US CLOUD Act

    Microsoft, OpenAI, AWS and Google are US corporations and are subject to the US CLOUD Act, even with data centres in Switzerland or the EU. This residual risk belongs in every risk assessment.

    Intelligence Index

    Performance: Artificial Analysis Intelligence Index, version 4.3.2, values as of 23 October 2026, scale 0 to 100. "ca." = estimated by Artificial Analysis. 128,000 tokens correspond to roughly 190 A4 pages. The traffic light is our risk assessment and does not replace legal advice. Data as of: 23 October 2026.

    Not listed: Meta Muse Wonder (Index 52) and Grok 4.7 by xAI. Both are among the strongest models but currently have no deployment route with processing in Switzerland or the EU.

    Comparison · Insurance brokers · Client data

    Three routes compared side by side.

    Choose up to three options. Data path, performance and cost sit next to each other, row by row.

    Microsoft Azure, Switzerland North region
    Processing
    e.g. Phoeniqs (kvant), Swisscom, Infomaniak, AlpineAI
    Processing
    Amazon Web Services, EU regions
    Processing
    Traffic light for Insurance brokers
    Green
    Green
    Yellow
    Processing and storage
    Switzerland
    Storage: Switzerland
    Switzerland
    Storage: Switzerland
    EU
    Storage: No prompt storage by Bedrock
    Best model on this route
    GPT-4ountil April 2027
    GLM 5.3 FlashPhoeniqs
    Claude Opus 5.5via EU inference
    Intelligence Index · scale 0 to 100
    ca. 8GPT-4o · estimated by Artificial Analysis
    42GLM 5.3 Flash
    58Claude Opus 5.5
    Document volume
    up to approx. 190 pagesat once in a single request
    up to approx. 1,500 pagesat once in a single request
    up to approx. 1,500 pagesat once in a single request
    Cost
    Low, per token, no base fee
    Medium, licence per user, sometimes plus usage
    Low to medium, per token
    What to watch for
    A Swiss data location satisfies the outsourcing rules most easily (Circular 2018/3 for insurers, applied analogously for brokers): choose the Standard deployment type (not Global or Data Zone), contract with confidentiality and audit rights. For new customers, GPT-4o (until April 2027) is effectively the only option; GPT-4.1 is deprecated.
    Well suited for property and commercial client data; for health data vet the provider in depth, hence yellow in the second rating.
    Like Data Zone EU: document EU processing, red for health data.
    Technical figures: coding index, tokens
    Coding index (Artificial Analysis)
    not yet measured
    72GLM 5.3 Flash
    78Claude Opus 5, strongest measured model of this option
    Context window in tokens
    128,000Tokens, i.e. word fragments
    1,000,000Tokens, i.e. word fragments
    1,000,000Tokens, i.e. word fragments
    What does the number mean?
    As of 23 October 2026

    Artificial Analysis Intelligence Index. Independent benchmark by Artificial Analysis. All models solve the same ten demanding task sets: agentic tasks and real office work (30%), general knowledge, long documents and hallucinations (30%), coding (20%) and scientific reasoning (20%). Scale 0 to 100, higher is better. The tests are deliberately very hard: the best model scores 58 today.

    Version 4.3.2, values as of 23 October 2026. "ca." = estimated by Artificial Analysis, the independent measurement is still pending.

    Your personal assessment

    We assess the result for your company personally.

    • Traffic light for all 13 options for your industry and your requirement
    • Recommended setup with data path and performance profile
    • Checklist for policy, data processing agreement and risk assessment
    Prefer to talk directly?Book an intro call

    Request your personal assessment

    Your selection is already filled in:

    Industry:
    Insurance brokers
    Data:
    Client data
    Requirement:
    Switzerland only
    No newsletter. You receive a confirmation and, if selected, notes on new models.

    What FINMA and data protection expect

    Independent insurance brokers are registered with FINMA and supervised directly, and supervisory notice 08/2024 applies to them too. Tied agents are supervised by the insurer itself: align AI use with the insurer in advance, its IT policy decides. Insurers face the full programme: Circular 2018/3 applies directly, and outsourcing of significant functions belongs in the business plan under Art. 4 ISA. The DPA applies to everyone, with heightened protection for health data.

    • AI inventory: which applications, which data, which model, which data location
    • Risk assessment per application, including cross-border processing and the US CLOUD Act
    • Policy for employees: approved tools, banned shadow AI, human oversight
    • Outsourcing contract with confidentiality, audit rights and data location (Circular 2018/3 directly for insurers, analogously for brokers)
    • For insurers: file the outsourcing of significant functions in the business plan with FINMA (Art. 4 ISA)
    • Data protection impact assessment for large-scale processing of health data (Art. 22 DPA)
    • Obtain explicit consent where the processing of health data relies on it (Art. 6(7) DPA)
    • Review cooperation agreements with insurers for data transfer and processing clauses
    • Client information in the policy terms or privacy policy reviewed

    This overview does not replace legal advice. Model availability and regions change constantly; October 2026 edition. Verify details with the provider before use.

    Sources

    We implement the recommended setup for you

    Humanyx accompanies you as a partner from choosing the operating route to ongoing operations: Switzerland only with Azure OpenAI Switzerland North, or the frontier model in the EU while your client data stays in Switzerland. Contract and governance are included. You receive finished AI employees for your processes, from document review to client reporting, instead of your own infrastructure project.

    • Processing and storage in Switzerland (Azure Switzerland North)
    • Contract, policies and documentation for data protection and supervision included
    • Fixed-price pilot, then ongoing operations by us