DATEV AI features vs. Salesforce Einstein / Agentforce: data protection per plan

not publicly documented

yes (concluded via online form with DATEV SmartCard/SmartLogin or by being stored in DATEV Einverständniserklärungen; effective 'upon confirmation of the conclusion of the contract by DATEV' (translated), DPA item 19 a)

yes (concluded via online form with DATEV SmartCard/SmartLogin or by being stored in DATEV Einverständniserklärungen; effective 'upon confirmation of the conclusion of the contract by DATEV' (translated), DPA item 19 a)

yes (on request)

yes (on request)

yes (on request)

yes (on request)

yes (on request)

yes (on request)

KI-Werkstatt page: 'Some prototypes in the DATEV KI-Werkstatt use the Microsoft Azure Open AI Services' (translated); inputs are 'not passed on to OpenAI and not used to train the models' (translated). No statement for the other prototypes; therefore not substantiated for the plan as a whole.

DATEV Copilot FAQ: 'Is my data used to train the AI? No, neither Microsoft nor OpenAI use your data for training.' (translated). Copilot is a DATEV cloud service under the DATEV service descriptions for the DATEV-Cloud-Lösungen Basis (doc. no. 0904407); unlike the service descriptions for Rechnungswesen, Steuern, Personalwirtschaft and Wissen, these contain no clause on cross-customer processing for 'Intelligente DATEV-Cloud-Lösungen'. Item 3.8, however, permits DATEV to process data 'for the purpose of developing new or further developing existing DATEV products in an appropriately secured environment' (translated); according to item 3.8 this does not apply if the customer concluded the DPA without the anonymisation agreement (deletion possible under DPA item 15). An express statement on DATEV's own model training with Copilot inputs is missing.

DATEV service descriptions for the DATEV-Cloud-Lösungen Rechnungswesen (accounting), Steuern (tax) and Personalwirtschaft (payroll), each item 1.8 'Service-related cross-customer data processing' (translated) (identical wording also for Wissen [knowledge]): for 'intelligent cloud solutions' (translated), the customer instructs DATEV to use the data sets 'set up by the customer for the „Intelligente DATEV-Cloud-Lösungen“' (translated) 'in the context of new and further development, quality assurance and operation' (translated); 'For machine learning or deep learning, this training forms the basis for the systems' learning' (translated); 'for example, data of different customers of the „DATEV-Cloud-Lösungen“ are analysed together' (translated); 'Personal data may still be processed in the resulting AI models as well' (translated); special categories of personal data may also be processed. The application of the AI models remains 'exclusively a DATEV-internal process' (translated). No possibility of objection is described in item 1.8; according to its wording, the optional deletion of the anonymisation agreement (DPA item 15b) only affects items 3.7/3.8, not item 1.8. The Basis service description (Copilot) contains no item 1.8 of this kind. Payroll benchmark uses 'real, anonymised DATEV payroll records' (translated); de-identification service: data is 'analysed in the DATEV data centre' (translated) and de-identified before AI processing.

Documentation does not differentiate by plan. Trust Layer: "prompts and generated responses are never stored or used to train the underlying third-party large language models" (Zero Data Retention). Einstein Platform Notices (Published 2026-05-22, q7): OpenAI, Azure, Google and Anthropic (via AWS) "have contracted to not otherwise retain or use Customer Data sent to the model" (apart from automated safety classification). Use by Salesforce itself: the Einstein Platform SPARC (Published 2026-07-24, q8) permits the use of customer data "on an aggregate basis for purposes such as research, marketing, analysis, and benchmarking, and other purposes reasonably required to develop, deliver, and provide ongoing innovation to the Covered Services", without personally identifiable data; the documentation contains no express statement on training Salesforce's own models. For Einstein Conversation Insights, Salesforce employees may review customer data "for the purposes of improving and training the product" (q7). SPARC and Notices apply to the "Covered Services branded as Agentforce" without differentiating by edition; Salesforce Foundations is not named separately there (q7, q8).

Documentation does not differentiate by plan. Trust Layer: "prompts and generated responses are never stored or used to train the underlying third-party large language models" (Zero Data Retention). Einstein Platform Notices (Published 2026-05-22, q7): OpenAI, Azure, Google and Anthropic (via AWS) "have contracted to not otherwise retain or use Customer Data sent to the model" (apart from automated safety classification). Use by Salesforce itself: the Einstein Platform SPARC (Published 2026-07-24, q8) permits the use of customer data "on an aggregate basis for purposes such as research, marketing, analysis, and benchmarking, and other purposes reasonably required to develop, deliver, and provide ongoing innovation to the Covered Services", without personally identifiable data; the documentation contains no express statement on training Salesforce's own models. For Einstein Conversation Insights, Salesforce employees may review customer data "for the purposes of improving and training the product" (q7).

Documentation does not differentiate by plan. Trust Layer: "prompts and generated responses are never stored or used to train the underlying third-party large language models" (Zero Data Retention). Einstein Platform Notices (Published 2026-05-22, q7): OpenAI, Azure, Google and Anthropic (via AWS) "have contracted to not otherwise retain or use Customer Data sent to the model" (apart from automated safety classification). Use by Salesforce itself: the Einstein Platform SPARC (Published 2026-07-24, q8) permits the use of customer data "on an aggregate basis for purposes such as research, marketing, analysis, and benchmarking, and other purposes reasonably required to develop, deliver, and provide ongoing innovation to the Covered Services", without personally identifiable data; the documentation contains no express statement on training Salesforce's own models. For Einstein Conversation Insights, Salesforce employees may review customer data "for the purposes of improving and training the product" (q7).

Documentation does not differentiate by plan. Trust Layer: "prompts and generated responses are never stored or used to train the underlying third-party large language models" (Zero Data Retention). Einstein Platform Notices (Published 2026-05-22, q7): OpenAI, Azure, Google and Anthropic (via AWS) "have contracted to not otherwise retain or use Customer Data sent to the model" (apart from automated safety classification). Use by Salesforce itself: the Einstein Platform SPARC (Published 2026-07-24, q8) permits the use of customer data "on an aggregate basis for purposes such as research, marketing, analysis, and benchmarking, and other purposes reasonably required to develop, deliver, and provide ongoing innovation to the Covered Services", without personally identifiable data; the documentation contains no express statement on training Salesforce's own models. For Einstein Conversation Insights, Salesforce employees may review customer data "for the purposes of improving and training the product" (q7).

Documentation does not differentiate by plan. Trust Layer: "prompts and generated responses are never stored or used to train the underlying third-party large language models" (Zero Data Retention). Einstein Platform Notices (Published 2026-05-22, q7): OpenAI, Azure, Google and Anthropic (via AWS) "have contracted to not otherwise retain or use Customer Data sent to the model" (apart from automated safety classification). Use by Salesforce itself: the Einstein Platform SPARC (Published 2026-07-24, q8) permits the use of customer data "on an aggregate basis for purposes such as research, marketing, analysis, and benchmarking, and other purposes reasonably required to develop, deliver, and provide ongoing innovation to the Covered Services", without personally identifiable data; the documentation contains no express statement on training Salesforce's own models. For Einstein Conversation Insights, Salesforce employees may review customer data "for the purposes of improving and training the product" (q7).

Documentation does not differentiate by plan. Trust Layer: "prompts and generated responses are never stored or used to train the underlying third-party large language models" (Zero Data Retention). Einstein Platform Notices (Published 2026-05-22, q7): OpenAI, Azure, Google and Anthropic (via AWS) "have contracted to not otherwise retain or use Customer Data sent to the model" (apart from automated safety classification). Use by Salesforce itself: the Einstein Platform SPARC (Published 2026-07-24, q8) permits the use of customer data "on an aggregate basis for purposes such as research, marketing, analysis, and benchmarking, and other purposes reasonably required to develop, deliver, and provide ongoing innovation to the Covered Services", without personally identifiable data; the documentation contains no express statement on training Salesforce's own models. For Einstein Conversation Insights, Salesforce employees may review customer data "for the purposes of improving and training the product" (q7).

Page: prototypes 'in a protected DATEV environment' (translated); for Azure OpenAI prototypes, inputs are 'sent as a request through the DATEV cloud-native data centre to Microsoft Azure' (translated) and processed there temporarily. Azure region not stated.

FAQ: 'DATEV Copilot uses Microsoft Azure services. The data you upload is processed in an area hosted by DATEV at Microsoft Azure in a European data centre.' (translated). Internet research (optional, permission 'Allow internet research' (translated)): according to the programme content (doc. no. 0904530), 'you transmit your data to Microsoft Corporation in the USA' (translated); 'Microsoft processes your data for its own purposes and on global servers' (translated); DATEV warns against transmitting confidential or legally protected information.

DATEV data centre/DATEV cloud according to the de-identification article and the AI overview. Help document 1000192 (as of 2026-09-18): Microsoft Ireland Operations Limited (IE) for 'DATEV Einspruchsgenerator, DATEV LEXchat AI functionalities' (translated) without transfer to a third country; transfers to a third country (USA) to Datadog, Inc. (DATEV Bankdatenservice) and Raintank, Inc./Grafana Labs (IT monitoring), among others, on the basis of the EU-U.S. DPF.

Infrastructure-and-Sub-processors documentation (Published 2026-10-02): Einstein Generative AI Services process at OpenAI in the USA (failover rerouting to Azure, priority to the org region), at Azure OpenAI in several regions incl. the EU (e.g. France, Germany, Sweden), at Google/AWS in the nearest disclosed region (e.g. Germany, Netherlands; failover via US endpoints); batch processing via Azure OpenAI uses global endpoints without geo-routing. According to a footnote, Agentforce Coworker only in certain regions (e.g. USA, Netherlands or Germany, France, Italy, Sweden). Org hosting on Hyperforce/AWS in selectable regions incl. Germany, France, Ireland and others.

Infrastructure-and-Sub-processors documentation (Published 2026-10-02): Einstein Generative AI Services process at OpenAI in the USA (failover rerouting to Azure, priority to the org region), at Azure OpenAI in several regions incl. the EU (e.g. France, Germany, Sweden), at Google/AWS in the nearest disclosed region (e.g. Germany, Netherlands; failover via US endpoints); batch processing via Azure OpenAI uses global endpoints without geo-routing. According to a footnote, Agentforce Coworker only in certain regions (e.g. USA, Netherlands or Germany, France, Italy, Sweden). Org hosting on Hyperforce/AWS in selectable regions incl. Germany, France, Ireland and others.

Infrastructure-and-Sub-processors documentation (Published 2026-10-02): Einstein Generative AI Services process at OpenAI in the USA (failover rerouting to Azure, priority to the org region), at Azure OpenAI in several regions incl. the EU (e.g. France, Germany, Sweden), at Google/AWS in the nearest disclosed region (e.g. Germany, Netherlands; failover via US endpoints); batch processing via Azure OpenAI uses global endpoints without geo-routing. According to a footnote, Agentforce Coworker only in certain regions (e.g. USA, Netherlands or Germany, France, Italy, Sweden). Org hosting on Hyperforce/AWS in selectable regions incl. Germany, France, Ireland and others.

Infrastructure-and-Sub-processors documentation (Published 2026-10-02): Einstein Generative AI Services process at OpenAI in the USA (failover rerouting to Azure, priority to the org region), at Azure OpenAI in several regions incl. the EU (e.g. France, Germany, Sweden), at Google/AWS in the nearest disclosed region (e.g. Germany, Netherlands; failover via US endpoints); batch processing via Azure OpenAI uses global endpoints without geo-routing. According to a footnote, Agentforce Coworker only in certain regions (e.g. USA, Netherlands or Germany, France, Italy, Sweden). Org hosting on Hyperforce/AWS in selectable regions incl. Germany, France, Ireland and others.

Infrastructure-and-Sub-processors documentation (Published 2026-10-02): Einstein Generative AI Services process at OpenAI in the USA (failover rerouting to Azure, priority to the org region), at Azure OpenAI in several regions incl. the EU (e.g. France, Germany, Sweden), at Google/AWS in the nearest disclosed region (e.g. Germany, Netherlands; failover via US endpoints); batch processing via Azure OpenAI uses global endpoints without geo-routing. According to a footnote, Agentforce Coworker only in certain regions (e.g. USA, Netherlands or Germany, France, Italy, Sweden). Org hosting on Hyperforce/AWS in selectable regions incl. Germany, France, Ireland and others.

Infrastructure-and-Sub-processors documentation (Published 2026-10-02): Einstein Generative AI Services process at OpenAI in the USA (failover rerouting to Azure, priority to the org region), at Azure OpenAI in several regions incl. the EU (e.g. France, Germany, Sweden), at Google/AWS in the nearest disclosed region (e.g. Germany, Netherlands; failover via US endpoints); batch processing via Azure OpenAI uses global endpoints without geo-routing. According to a footnote, Agentforce Coworker only in certain regions (e.g. USA, Netherlands or Germany, France, Italy, Sweden). Org hosting on Hyperforce/AWS in selectable regions incl. Germany, France, Ireland and others.

• Microsoft Azure OpenAI Services (some prototypes)

• Microsoft Azure (AI services, European data centre, according to the Copilot FAQ)
• OpenAI models via Azure (according to the FAQ, OpenAI does not use data for training)
• Microsoft Bing (only with internet research enabled; independent processing by Microsoft)
• Service descriptions Basis item 3.4: 'DATEV Copilot: Microsoft Ireland Operations Limited' (Dublin, IE), purpose 'AI functions' (translated); otherwise help document 1000192

• Microsoft Ireland Operations Limited (Einspruchsgenerator, LEXchat AI functionalities)
• Amazon Web Services EMEA SARL (operation/hosting of DATEV products)
• Microsoft Deutschland GmbH (DATEV products, data centre services)
• Google Ireland Ltd. (DATEV platform, chatbot use)
• Datadog, Inc. (USA, DATEV Bankdatenservice, DPF)
• others according to help document 1000192

• OpenAI, L.L.C. (generative AI, USA; on failover rerouting to an Azure endpoint)
• Microsoft Corporation (Microsoft Azure / Azure OpenAI; regions incl. USA, Switzerland, France, Germany, Sweden, UK and others)
• Google LLC (generative AI; "disclosed region closest to Customer's Salesforce Org", e.g. Germany, Netherlands; failover via Google endpoint in the USA)
• Amazon Web Services, Inc. (hosting and generative AI provider, including Anthropic models via AWS according to the Notices q7; Agentforce Speech: AWS Polly/Transcribe)
• Bright Data Inc. (web search for generative AI, only if enabled by the customer)
• SuSea, Inc. (dba You.com; web search for generative AI, only if enabled by the customer)
• Temporal Technologies Inc. (orchestration of background workflows; incl. USA, Germany, UK)
• Eleven Labs Inc. (text-to-speech/speech-to-text, Agentforce Speech Foundations)
• Deepgram, Inc. (speech-to-text, Agentforce Speech Foundations; also self-hosted on AWS)

• OpenAI, L.L.C. (generative AI, USA; on failover rerouting to an Azure endpoint)
• Microsoft Corporation (Microsoft Azure / Azure OpenAI; regions incl. USA, Switzerland, France, Germany, Sweden, UK and others)
• Google LLC (generative AI; "disclosed region closest to Customer's Salesforce Org", e.g. Germany, Netherlands; failover via Google endpoint in the USA)
• Amazon Web Services, Inc. (hosting and generative AI provider, including Anthropic models via AWS according to the Notices q7; Agentforce Speech: AWS Polly/Transcribe)
• Bright Data Inc. (web search for generative AI, only if enabled by the customer)
• SuSea, Inc. (dba You.com; web search for generative AI, only if enabled by the customer)
• Temporal Technologies Inc. (orchestration of background workflows; incl. USA, Germany, UK)
• Eleven Labs Inc. (text-to-speech/speech-to-text, Agentforce Speech Foundations)
• Deepgram, Inc. (speech-to-text, Agentforce Speech Foundations; also self-hosted on AWS)

• OpenAI, L.L.C. (generative AI, USA; on failover rerouting to an Azure endpoint)
• Microsoft Corporation (Microsoft Azure / Azure OpenAI; regions incl. USA, Switzerland, France, Germany, Sweden, UK and others)
• Google LLC (generative AI; "disclosed region closest to Customer's Salesforce Org", e.g. Germany, Netherlands; failover via Google endpoint in the USA)
• Amazon Web Services, Inc. (hosting and generative AI provider, including Anthropic models via AWS according to the Notices q7; Agentforce Speech: AWS Polly/Transcribe)
• Bright Data Inc. (web search for generative AI, only if enabled by the customer)
• SuSea, Inc. (dba You.com; web search for generative AI, only if enabled by the customer)
• Temporal Technologies Inc. (orchestration of background workflows; incl. USA, Germany, UK)
• Eleven Labs Inc. (text-to-speech/speech-to-text, Agentforce Speech Foundations)
• Deepgram, Inc. (speech-to-text, Agentforce Speech Foundations; also self-hosted on AWS)

• OpenAI, L.L.C. (generative AI, USA; on failover rerouting to an Azure endpoint)
• Microsoft Corporation (Microsoft Azure / Azure OpenAI; regions incl. USA, Switzerland, France, Germany, Sweden, UK and others)
• Google LLC (generative AI; "disclosed region closest to Customer's Salesforce Org", e.g. Germany, Netherlands; failover via Google endpoint in the USA)
• Amazon Web Services, Inc. (hosting and generative AI provider, including Anthropic models via AWS according to the Notices q7; Agentforce Speech: AWS Polly/Transcribe)
• Bright Data Inc. (web search for generative AI, only if enabled by the customer)
• SuSea, Inc. (dba You.com; web search for generative AI, only if enabled by the customer)
• Temporal Technologies Inc. (orchestration of background workflows; incl. USA, Germany, UK)
• Eleven Labs Inc. (text-to-speech/speech-to-text, Agentforce Speech Foundations)
• Deepgram, Inc. (speech-to-text, Agentforce Speech Foundations; also self-hosted on AWS)

• OpenAI, L.L.C. (generative AI, USA; on failover rerouting to an Azure endpoint)
• Microsoft Corporation (Microsoft Azure / Azure OpenAI; regions incl. USA, Switzerland, France, Germany, Sweden, UK and others)
• Google LLC (generative AI; "disclosed region closest to Customer's Salesforce Org", e.g. Germany, Netherlands; failover via Google endpoint in the USA)
• Amazon Web Services, Inc. (hosting and generative AI provider, including Anthropic models via AWS according to the Notices q7; Agentforce Speech: AWS Polly/Transcribe)
• Bright Data Inc. (web search for generative AI, only if enabled by the customer)
• SuSea, Inc. (dba You.com; web search for generative AI, only if enabled by the customer)
• Temporal Technologies Inc. (orchestration of background workflows; incl. USA, Germany, UK)
• Eleven Labs Inc. (text-to-speech/speech-to-text, Agentforce Speech Foundations)
• Deepgram, Inc. (speech-to-text, Agentforce Speech Foundations; also self-hosted on AWS)

• OpenAI, L.L.C. (generative AI, USA; on failover rerouting to an Azure endpoint)
• Microsoft Corporation (Microsoft Azure / Azure OpenAI; regions incl. USA, Switzerland, France, Germany, Sweden, UK and others)
• Google LLC (generative AI; "disclosed region closest to Customer's Salesforce Org", e.g. Germany, Netherlands; failover via Google endpoint in the USA)
• Amazon Web Services, Inc. (hosting and generative AI provider, including Anthropic models via AWS according to the Notices q7; Agentforce Speech: AWS Polly/Transcribe)
• Bright Data Inc. (web search for generative AI, only if enabled by the customer)
• SuSea, Inc. (dba You.com; web search for generative AI, only if enabled by the customer)
• Temporal Technologies Inc. (orchestration of background workflows; incl. USA, Germany, UK)
• Eleven Labs Inc. (text-to-speech/speech-to-text, Agentforce Speech Foundations)
• Deepgram, Inc. (speech-to-text, Agentforce Speech Foundations; also self-hosted on AWS)

'The platform can be used free of charge' (translated); login via DATEV account and SmartCard/SmartLogin (as of 2026-10-18)

DATEV shop (item no. 90431): 'The prerequisite is ordering a free licence via this shop page' (translated); ordering also via the MyDATEV ordering and set-up assistant. Usage limit: rolling 2-hour window, if exceeded blocked for 2 hours (doc. no. 1049806). MyDATEV Kanzlei and LEXinform plus require a licence (as of 2026-10-18)

Billing according to the service descriptions 'in accordance with the current DATEV price list' (translated); the DATEV price list for firms is offered as a PDF download and was not evaluated (as of 2026-10-18)

USD 0 (Salesforce Foundations, incl. Agentforce Builder, Prompt Builder, Agentforce Coworker and others; as of 2026-10-14)

USD 500 per 100,000 Flex Credits (as of 2026-10-14; pricing page shows a USD/EUR converter, EUR 500 per 100k)

USD 2 per conversation; for customer-facing agents, cannot be used in the same org as Flex Credits (as of 2026-10-14)

USD 5 per user/month, requires Flex Credits (as of 2026-10-14)

USD 125 per user/month (Sales/Service/Field Service/Operations add-ons) or USD 150 per user/month (Industries add-ons), unlimited employee agent usage (as of 2026-10-14)

from USD 550 per user/month, includes Agentforce add-on and 2.75 million Flex Credits per org/year (as of 2026-10-14). Sales Cloud pricing page (as of 2026-10-18): Max USD 550 or EUR 550 per user/month, billed annually, no indication whether VAT is included (q9)

yes

yes

yes

yes

yes

yes

yes

yes

yes