1. Scope and availability
This Addendum supplements the Agreement with PCGROUP, LLC / JBComm for AIMarket / OmniReach AI functions actually enabled and contracted. Master Terms definitions apply. Public descriptions do not promise release or availability of every function. DPA and AUP duties continue.
2. Assistive role, permissions, and actions
AI is assistive unless an action-capable function is expressly enabled and authorized under the Agreement and appropriate tenant policies. Data access and actions remain subject to assigned user permissions. This clause does not independently authorize external communications, account changes, or financial actions.
You determine authorized actions, permission scope, and appropriate review, and remain responsible for actions you authorize your users or enabled AI functions to take. We retain our own legal/contract duties. Do not seek other tenants’ information or bypass access limits.
3. AI Output and appropriate review
AI Output may be inaccurate, incomplete, biased, misleading, unsuitable, or non-unique. Apply appropriate human judgment and verify relevant facts, recipients, amounts, permissions, and legal requirements before reliance or action. Similar output gives no rights in another tenant’s information.
AI Output is not legal, financial, tax, or other regulated professional advice. No correctness, uniqueness, lawful suitability, business outcome, or revenue is guaranteed. Consequential and customer-facing uses require suitable review.
4. Data and AI Output rights
The Customer retains Customer Data, Content, and Tenant Intelligence. As between Customer and PCGROUP, any rights we hold in AI Output generated specifically for that Customer are assigned to it to the extent legally transferable, excluding underlying platform and third-party materials. No copyrightability, exclusivity, originality, or source rights we do not hold are guaranteed.
AI summaries/inferences about identifiable people or your business do not become our Derived Data merely because AI created them. Prospect and Enriched Data remain subject to source and third-party rights.
5. No shared-model training without separate authorization
We will not use identifiable Customer Data, Customer Content, Prospect Data, or tenant-specific data to train a shared or general model for unrelated customers unless the Customer separately authorizes that use.
Separate authorization must be informed, specific to data/purpose/recipients, recorded independently from general acceptance, and granted by someone with Customer authority. Applicable law and underlying rights must permit it. A general improvement clause, feedback request, or integration setting is not authorization. Where the DPA applies, an authorization also requires lawful documented instructions and a reviewed processing schedule; no existing annex is silently expanded.
The restriction also applies to model/service providers processing such information on our behalf. Require consistent lawful terms or refrain from affected use. Independent licensed sources remain subject to their own permissions; Customers cannot grant rights they do not hold.
6. Tenant context and permitted improvement
Authorized Customer Data may support that Customer’s enabled functions, including tenant-specific context, memory, and derived intelligence, only within limited service purposes, permissions, instructions, and retention rules. These are purpose statements, not disclosure of internal AI methods.
Improvement must follow the Agreement/law and must not expose one tenant’s identifiable information to another. Derived Data requires the Master Terms’ narrow conditions and protection against re-identification. Pseudonymous or customer-specific data is not presumed anonymous. This clause does not override section 5 or state service-provider restrictions.
7. Providers, security, and abuse prevention
AI features may use third-party model/service providers. Processing for us is subject to appropriate contracts, applicable DPA duties, and approved Subprocessor disclosures. Necessary security, maintenance, support, and abuse prevention must remain within lawful role/purpose limits.
An abuse-prevention label does not create unlimited retention, unrelated training permission, or cross-tenant disclosure. Access, incidents, deletion, and security obligations continue. Independently selected Customer integrations may have distinct terms requiring separate review.
8. Automated actions and external communications
Enabled actions require lawful authorization and review/safeguards suitable to risk. Before sending or authorizing AI-generated communications, review purpose, content, identity, recipients, permissions, consent, opt-outs, and required disclosures. Automation does not create consent or replace channel duties.
Make AI/automated-interaction and recording disclosures where law or applicable provider policies require. Do not deceptively imply human production where the implication is misleading or disclosure required. Payment/account actions require lawful authorization under applicable service terms.
9. High-impact and sensitive-use restrictions
You must not use AIMarket AI to make fully automated decisions having legal or similarly significant effects concerning employment, credit, housing, insurance, healthcare eligibility, or similarly regulated high-impact matters unless the use is specifically supported, lawful, expressly authorized in a signed agreement, and subject to appropriate safeguards and meaningful human oversight.
Ordinary subscriptions to Prospect, VCC, Portfolio, or CSSM do not authorize these uses. Do not use AI for unlawful discrimination, surveillance, stalking, sensitive-person targeting, or profiling children. Sensitive-use limits in the AUP and service terms apply. Any permitted exception needs documented scope and safeguards.
10. Conflicts and continuing obligations
This Addendum controls AI/data-use subject matter without overriding stricter lawful DPA or negotiated protections. Lower-ranked documentation or generic service improvement cannot silently permit prohibited shared-model training. Data-use restrictions continue while relevant data is processed or retained.
Amendments and separate authorizations follow the Agreement/law. Existing Master Terms liability and indemnity treatment remains, subject to the DPA’s mandatory-law provisions; no new cap or provider indemnity is created.