Humanyx
    AI comparison · Oct. 2026

    Which AI may wealth managers 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.

    IndustryWealth managers and financial service providers

    Which data should go into the AI?

    Health data appears as soon as you choose insurance brokers.

    Where may your data be processed?

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

    For Wealth managers and financial service providers

    Our recommendation: client data only in Switzerland. Processing abroad is only permitted with a risk assessment, a contract and protection of the client data, and a breach of professional secrecy is a criminal offence (Art. 69 FinIA).

    Your result · Wealth managers · 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 wealth managers and financial service providers

    Our recommendation: client data only in Switzerland. Processing abroad is only permitted with a risk assessment, a contract and protection of the client data (see the Swiss Bankers Association's Cloud Guidelines), and a breach of professional secrecy is a criminal offence (Art. 69 FinIA).

    Recommended for wealth managers

    Strict: client data only in Switzerland

    • Policy for all staff: client data only via the Swiss pipeline
    • Swiss AI platform: working surface for all, external models disabled
    • Azure OpenAI Switzerland: for automations such as document review or reports
    In addition, for daily office work

    Copilot only for work without client data

    • Processing in the EU: so only for drafts, research and internal documents
    • Flex Routing off: otherwise Copilot also processes outside Europe at peak load
    • In-country storage on: only then does Copilot store data in Switzerland
    When frontier models are needed

    Reference architecture: frontier models with client data in Switzerland

    • Frontier model in the EU: receives only pseudonymised content
    • Mapping stays in Switzerland: the model never sees names or account numbers
    • Secure with care: risk assessment, contract and client information
    How the reference architecture works
    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

    • 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.

    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.

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

    Sources: FDPIC: Cross-border transfer of personal dataMicrosoft Learn: Region availability for Foundry Models sold by AzureMicrosoft Learn: Deployment types in Microsoft Foundry Models (Global, Data Zone, Standard, Provisioned)Microsoft Learn: Data residency for Microsoft 365 CopilotMicrosoft Learn: «Store Microsoft 365 customer data in-country» settingMicrosoft Learn: Flex routing (EU and EFTA)AWS: Model support by AWS Region in Amazon BedrockGoogle Cloud: Data residency for generative AIFINMA: Supervisory notice 08/2024 on governance and risk management when using AIFINMA: Circular 2018/3 OutsourcingFedlex: Financial Institutions Ordinance, Art. 17 (outsourcing)Fedlex: Financial Institutions Act, Art. 69 (professional secrecy)Fedlex: Banking Act, Art. 47 (banking secrecy)Phoeniqs: Active ModelsPhoeniqs: Preview Models (Kimi K3, GLM 5.3, DeepSeek V4.1 Flash)Phoeniqs: Release note on the model portfolio, May 2026Artificial Analysis: Intelligence Index (methodology and values, v4.3.2)Artificial Analysis: GPT-4o (value estimated)Artificial Analysis: GPT-5.1 (value estimated)Artificial Analysis: gpt-oss-120bIndex values of all models (modelgrep, data from Artificial Analysis)Infomaniak: open-source modelsSwisscom Swiss AI Platform: model overviewAlpineAI: SwissGPTMicrosoft Learn: region availability of Azure OpenAI models (Data Zone EU)OpenAI: GPT-6 and EU data residency in the APIOpenAI: data and inference residency for ChatGPT EnterpriseGoogle Cloud: data residency of models on Vertex AIMeta Model API: Geographic Use Policy (Muse Wonder)

    Comparison · Wealth managers · 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 Wealth managers
    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
    The reference setup under FINMA logic: Swiss data location, choose the Standard deployment type (not Global or Data Zone), outsourcing contract under Art. 17 FinIO or Circular 2018/3 with confidentiality and audit rights. For new customers, GPT-4o (until April 2027) is effectively the only option. Apply for Modified Abuse Monitoring: without this approval Microsoft stores flagged samples, and authorised staff can review them even in the Swiss region.
    The simplest FINMA-compliant surface without US ties: check ISO 27001, references and confidentiality, disable optional US models.
    Like Azure Data Zone EU; also cover AWS as a sub-processor in the contract. AWS evidences the FINMA requirements with a dedicated ISAE 3000 Type 2 report.
    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:
    Wealth managers
    Data:
    Client data
    Requirement:
    Switzerland only
    No newsletter. You receive a confirmation and, if selected, notes on new models.

    What data protection requires

    Cloud AI is usually commissioned processing under the DPA, and responsibility stays with you.

    • Commissioned processing: contract with confidentiality, audit rights and data location (Art. 9 DPA)
    • Data abroad: only with adequate protection, safeguards or an exception (Art. 16 and 17 DPA)
    • Policy: define approved tools, ban shadow AI
    • Transparency: keep the privacy policy and records of processing up to date (Art. 12 and 19 DPA)

    What the supervisor additionally requires

    Even when outsourcing to the cloud, you remain fully responsible to your supervisor and your clients.

    • Professional secrecy: a breach is a criminal offence, even through negligence (Art. 69 FinIA)
    • Outsourcing: document it, notify significant outsourcing to your supervisory organisation (Art. 17 FinIO)
    • AI governance: inventory, risk assessment and policies (FINMA supervisory notice 08/2024)
    • Contract: confidentiality and audit rights for your supervisory organisation and FINMA (Circular 2018/3, applied analogously)

    We implement the recommended setup for you

    Humanyx accompanies you as a partner from choosing the operating route to ongoing operations of your Swiss pipeline. Contract and governance are included. You get finished AI employees for your processes instead of an infrastructure project.

    • Switzerland only: processing and storage in Switzerland
    • Included: contract, policies and documentation for supervision and data protection
    • Fixed-price pilot: then ongoing operations by us