Platforms
What ChatGPT is, how the consumer, Team and Enterprise tiers differ, what businesses realistically use it for, and the controls required before staff use it with company information.
8 minute read · Written by the Fresh Mango AI team
ChatGPT is a conversational AI assistant developed by OpenAI, built on its GPT family of large language models. Launched publicly in November 2022, it lets users hold a natural-language conversation with a model that can draft, summarise, analyse, translate, reason through problems, write code, browse the web and work with uploaded files. Business and enterprise tiers add administrative control and contractual data protections.
OpenAI released ChatGPT as a research preview in November 2022. Within two months it had reached an estimated hundred million users, the fastest consumer software adoption then recorded. The technology underneath was not new — instruction-tuned transformer models had existed for several years — but the conversational interface removed every barrier to use. Anyone who could type could now direct a capable language model.
For businesses this created an unusual dynamic. The technology entered organisations from the bottom up, through individual employees using personal accounts, months or years before IT and compliance had formed a position on it. Most of the governance work we now do exists because of that sequence. The practical question in almost every organisation is not whether staff use ChatGPT, but whether they do so on managed accounts with appropriate protections.
Individual accounts. Conversations may be used to improve models unless the user opts out, there is no central administration, and no organisational audit trail. Not appropriate for confidential business content.
Designed for small groups. Business data is excluded from model training by default, with a shared workspace, shared custom GPTs and basic admin controls.
Adds SSO, domain verification, SCIM provisioning, compliance API access, longer context, and enterprise-grade security commitments including SOC 2 reporting. Business data is not used to train models.
For embedding models directly into your own applications and workflows. Azure OpenAI additionally keeps the deployment inside your Azure tenancy and regional boundary, which often decides the matter for regulated clients.
The dominant patterns are drafting and rewriting (proposals, bids, marketing copy, policies, job descriptions), analysis of pasted or uploaded material (contracts, transcripts, spreadsheets, research), structured thinking (options appraisals, risk registers, project plans), technical work (formulas, scripts, data transformation, code review) and translation. Custom GPTs extend this by packaging a persistent instruction set and reference documents into a reusable assistant — a bid-writing assistant loaded with your win themes, or a support assistant loaded with your product documentation.
ChatGPT's strengths relative to an in-suite assistant like Microsoft 365 Copilot are breadth, flexibility with uploaded material, and the ability to sustain a long analytical conversation. Its limitation in a Microsoft-centric business is that it does not have native, permissioned access to your tenant content unless you deliberately connect it. Many of our clients run both: Copilot for work inside documents and meetings, ChatGPT for open-ended analysis and content production.
Provide Team or Enterprise accounts to the roles that need them. Prohibition without provision reliably produces shadow usage on personal accounts.
State plainly what must never be entered: personal data beyond a defined threshold, client-confidential material without consent, credentials, unpublished financials, safeguarding or health information.
Any factual claim, citation, figure or legal reference must be checked against a primary source before it is relied upon or sent externally.
The person who sends the output owns the output. Make that explicit in your AI usage policy.
There is no universally correct answer, and the differences narrow with each model release. In practice we recommend ChatGPT Enterprise or Team where the work is open-ended content production and analysis across many file types; Microsoft 365 Copilot where the value is in grounded access to existing organisational content; Claude where long-document accuracy and careful reasoning dominate; and Gemini where the organisation runs on Google Workspace. The decision should follow your data estate and your regulatory position rather than model benchmarks, which change monthly.
OpenAI
OpenAI
Microsoft Learn
Book a free 30-minute discovery session with an AI consultant, or request information and we will send the detail you need within one business day.
Recommended · Free · 30 minutes
Opens our live booking calendar in a new tab — pick any slot that suits you.
Prefer not to book yet?
Send a short enquiry and we will come back with the detail you need.
FAQs
There is a free tier, but it lacks administrative control, audit and the contractual data protections most organisations require. For business use, ChatGPT Team or Enterprise is the appropriate starting point.
OpenAI states that business data submitted through Team, Enterprise and the API is not used to train its models by default. Consumer tiers behave differently, which is why sanctioned accounts matter.
A saved configuration of ChatGPT with its own instructions, tone, reference documents and optional tool connections. It lets a team reuse a well-designed assistant — for example for bid writing or first-line support — without writing code.
Not by default. Connections can be built through connectors or the API, with permissions carefully scoped. If grounded access to Microsoft 365 content is the primary requirement, Microsoft 365 Copilot is usually the more direct route.
Highly capable on language tasks, and unreliable on specific facts that are not in the conversation. It can produce plausible but invented references. Treat outputs as drafts requiring verification, particularly for legal, financial, medical or regulatory content.
Related articles
Explore the connected topics in the Fresh Mango AI Knowledge Centre.
Platforms
Claude is a family of large language models and an AI assistant built by Anthropic, an AI safety company founded in 2021. Claude is designed around a …
Read this guideFoundations
Generative AI is a class of artificial intelligence that creates new content — text, images, code, audio or video — in response to a prompt, rather th…
Read this guideGovernance
An AI usage policy is a short internal document that tells employees which AI tools are approved, what information may and may not be entered into the…
Read this guideBrowse the full AI Knowledge Centre, see sector guidance in industries, or review measured results in our case studies.
Decision brief
Straight answers to the five questions that decide whether an AI project is worth starting.
Fresh Mango AI is the artificial intelligence practice of Fresh Mango Technologies, an IT, cyber security and cloud provider that has supported businesses since 2004 from offices in Ripon, Leeds, Skipton and Tortola in the British Virgin Islands. The same engineers who secure your identity, data and Microsoft 365 tenant advise on your AI adoption, so recommendations are grounded in what your estate can actually support rather than in vendor marketing.
Clients typically release several hours per person per week on drafting, summarising, searching and reporting, shorten document and approval cycle times, and remove manual re-keying between systems. Every engagement starts by baselining the work involved so that the benefit is measured in hours released and cycle time reduced, not in licences purchased.
Doing nothing is not a neutral position. Staff adopt consumer AI tools on their own, so company and client data leaves your control without record; competitors compress the cost of proposals, reporting and service delivery; and permission sprawl inside your file estate remains unaddressed, which becomes an incident the moment AI search is switched on. Delay also compounds the UK GDPR and EU AI Act governance work that will eventually be required of you anyway.
You book a free 30-minute discovery session. We ask about your objectives, systems and constraints, tell you honestly whether AI is the right answer, and set out a recommended first step — usually an AI Productivity & Readiness Assessment or an AI Governance & Security Project. You receive a written summary and a proposal only if there is a clear case for one.
Book a free 30-minute discovery session with an AI consultant using our live booking calendar, or request information and we will reply within one business day. You can also call the UK office on +44 (0) 1765 606700 or the BVI office on +1 (284) 340 0466.
Recommended next step
Free, 30 minutes, no obligation — speak directly with an AI consultant about your objectives and the fastest safe route to results.
Book a Free 30-Minute Discovery SessionRequest InformationTalk it through with an AI consultant, or request written information — whichever suits you.
Your next step
Whether you're exploring AI for the first time or looking to scale existing AI initiatives, our specialists can help you identify practical opportunities and avoid common pitfalls.
