Lead360 AI Product Terms

Covered features

These AI Product Terms apply to Ask360 bots, Piece Studio, Memory Brain, Sales360, content generation, summaries, recommendations, classification, scoring, model routing, website generation, automated responses, and any feature identified as AI, model-assisted, predictive, generative, or automated.

Clear disclosure to people

A customer must clearly tell a person when that person is interacting with a bot or when a material communication is generated or handled automatically. The disclosure must be understandable, timely, and appropriate to the channel. A customer must not instruct a Lead360 bot to deny that it is automated, impersonate a real person without authorization, or create a false impression of live human review.

Where law or context requires it, the customer must also explain how a person can reach a human, object, correct information, withdraw consent, or stop further messages.

Inputs, permissions, and sensitive data

The customer is responsible for AI inputs, connected sources, prompts, files, websites, retrieval content, Memory Sets, Piece instructions, and the rights needed to process them. Do not submit passwords, private keys, payment authentication data, government identifiers, sensitive personal information, or confidential third-party content unless the use is necessary, lawful, and specifically approved.

Standard accounts must not use AI features for PHI. A written health-data approval and applicable agreements are required before any regulated health data is processed.

Model and provider processing

Lead360 may use locally hosted models or approved model and observability providers to complete a request. Current configured services may include Cerebras for inference and LangSmith for tracing and diagnostics. Selected inputs, retrieved context, outputs, trace identifiers, evaluations, and technical metadata may be processed by the provider needed for the feature.

Provider availability, retention, and use are governed by Lead360 provider configuration and applicable provider terms. A customer that requires a particular model route, region, retention setting, or no-external-processing configuration must obtain written confirmation before submitting restricted data.

Outputs and human review

Outputs are probabilistic and may be inaccurate, incomplete, outdated, biased, unsafe, or unsuitable. The customer must evaluate accuracy, tone, legality, rights, and context before relying on or sending an output. Lead360 does not guarantee that a model will follow every instruction or that the same input will produce the same output.

• Use qualified human review for material customer, legal, health, employment, credit, housing, insurance, or financial content.
• Verify names, dates, prices, claims, citations, permissions, consent, and commitments.
• Do not present generated content as professional advice unless an appropriately qualified person has reviewed and accepted it.
• Provide a correction and escalation path for people affected by an output.

Consequential decisions

Lead360 AI features must not be the sole basis for decisions that determine a person eligibility, access, rights, safety, health treatment, employment, housing, credit, insurance, education, immigration, legal services, or similarly significant outcome. Where automated support is legally permitted, the customer must conduct an impact assessment, test for accuracy and unfair bias, provide meaningful human review, maintain an appeal path, and give required notices.

Sales360 research and outreach

Public-source business data and generated inferences may be wrong or may relate to a person. Customers must verify source authorization, lawful basis, accuracy, relevance, marketing consent, and do-not-contact status before outreach. A fit score, pain point, or recommendation is not a fact and must not be represented as one.

Lead360 and customers must honour objections and suppression requests. Suppressed records may be retained in minimal form to prevent re-collection.

Memory, learning, and deletion

Memory Brain stores reviewed business context. Memory Sets bind approved revisions to a Piece. Learning Collections are saved filters, and Live Insights are verified aggregate findings. These are different functions. A retrieval window controls what content is used for a response, but does not necessarily delete source records, vectors, exports, traces, or backups.

Customers should remove outdated or unauthorized sources, publish corrected revisions, review Piece bindings, and request verified erasure when deletion is legally required.

Prohibited AI uses

• Deceptive impersonation, undisclosed bots, fabricated endorsements, or false evidence.
• Illegal surveillance, stalking, harassment, discrimination, or exploitation.
• Generation or distribution of malware, credential theft, evasion tools, or instructions intended to cause harm.
• Unlawful biometric identification, sensitive-trait inference, or profiling.
• Manipulation of children or vulnerable people, sexual exploitation, or non-consensual intimate content.
• Unreviewed professional advice or automated consequential decisions.
• Attempts to bypass model, tenant, tool, approval, rate, scope, or security controls.

Testing, monitoring, and enforcement

Customers must test bots, Pieces, tools, scopes, destinations, fallback behavior, and escalation before publication. Lead360 may log, evaluate, rate-limit, block, or require confirmation for AI activity to secure the service and enforce these terms. Lead360 may suspend an AI feature where it presents legal, security, safety, provider, or reputational risk.

Rights and responsibility

Rights in Customer Data remain with the customer. Rights in outputs may depend on law, provider terms, and third-party material, and Lead360 does not guarantee exclusivity or copyright protection. The customer is responsible for its use, editing, publication, and distribution of outputs. The disclaimers, indemnity, and liability limits in the Lead360 Terms apply.