No guaranteed outcomes
AiUse builds systems, workflows, pages, AI agents and automation logic. We do not guarantee rankings, revenue, ROI, lead volume, deliverability or sales results. Outcomes depend on offer quality, market, budget, data, team execution and external platforms.
AI implementation boundaries
AI should support business workflows, not make high-risk or regulated decisions without human review. We recommend fallback rules, logs, monitoring, clear responsibility and manual approval for sensitive contexts.
Data access
Client access should follow minimum necessary access, role-based permissions, secure credential sharing and documented handoff. Do not send passwords through public chat or unapproved channels.
GDPR-aware delivery
For EU and UK contexts, projects should consider data minimization, lawful basis, processor relationships, retention, access rights, international transfers and records of processing. AiUse can help design privacy-aware workflows, but final legal compliance belongs to the client and counsel.
US privacy considerations
US clients may have obligations under state privacy laws, sector rules, platform policies and email or messaging regulations. Outreach and automation must respect opt-out handling, suppression lists and deliverability rules.
Upwork and NDA projects
AiUse can work with Upwork-style scopes, milestones and documentation. NDA terms can be reviewed before a project starts. Scope, acceptance criteria, data access and deliverables should be explicit before implementation.
Third-party platforms
Integrations may depend on APIs, rate limits, paid plans and platform policy changes from CRMs, AI providers, email tools, messaging platforms and integration platforms. Final stack decisions depend on the client's current tools and privacy requirements.
Who else touches your data
We name our subprocessors rather than referring to trusted partners. Google supplies analytics, mail and calendar booking; the site is served by a commercial hosting provider; the AI agent runs on a commercial model API. Where a client system is involved — a CRM, a mailbox, an ad account — we work inside your tenancy under credentials you issue and can revoke, and we do not copy data out of it except where a written scope says so.
Who owns what we build
You do. Accounts, code, prompts, CRM structure, documentation and any content produced during the engagement belong to the client. There is no proprietary platform to be locked into and no licence that expires when the invoice stops. If we stop working together, the practical handover is a list of credentials and a document explaining how the system is put together — not a negotiation.
What happens when an engagement ends
We remove our access within five working days of the final invoice being settled, or immediately on request. Working copies of client data held for the project are deleted within 30 days, except where an invoice or contract must be retained for accounting purposes. You can ask for confirmation in writing and we will send it.
AI regulation and high-risk use
The EU AI Act phases in obligations that depend on how a system is used rather than on the model behind it. Everything we build is intended for low-risk commercial use: answering enquiries, qualifying leads, drafting text for human review, moving records between systems. We do not build systems that make or materially influence decisions about employment, credit, insurance, education access, or any other category the Act treats as high risk, and we will say so rather than take the work. Where a client wants automated decision-making with legal or similarly significant effects on individuals, that requires their own legal assessment before anything is built.
Accuracy, and what a model can and cannot be trusted with
Language models produce fluent output regardless of whether they know the answer. Every system we deliver is built with that assumption: binding statements — prices, delivery dates, technical specifications — are escalated to a person by rule rather than generated, sources are constrained to documents the client supplies, and there is always a path to a human. Where a client asks for an agent that quotes prices autonomously, we will explain what can go wrong and require it to be agreed in writing.
Marketing claims we do not make
We do not promise rankings, lead counts, revenue figures, ROAS, deliverability rates or a position in an AI answer. Anyone who does is either guessing or describing branded queries you already own. What is contractual is scope, deliverables and dates; what is reported is measurement, including when it says the work is not paying for itself.
Contact
For legal or project-scope questions, contact go@aiuse.agency.