Find the service boundary
We record the canonical domain, redirects, registration history, archived pages and contact routes. Advertising links and lookalike pages are traced rather than assumed to belong to the service.
Dubaicat starts with the service itself: the domain a customer reaches, the company named in the terms, the task described as automated, and the parties that receive data or money. Records that agree, records that conflict and facts we could not obtain stay separate.
The same page holds the assessment rules and the publication list. It does not certify a service as safe, legitimate or profitable.
No assessments have been published as of .
The file follows a customer journey from the first domain visit through account closure or a withdrawal request. A step is reported only when the cited record or observed screen supports it.
We record the canonical domain, redirects, registration history, archived pages and contact routes. Advertising links and lookalike pages are traced rather than assumed to belong to the service.
Names in the terms, privacy notice, payment screen and support correspondence are matched to company and regulator records for the relevant country. Incorporation proves that an entity exists; it does not establish permission to provide a financial service.
Research, alerts, copy trading, rule-based orders, managed accounts and referrals are different offers. We identify the task performed, where execution occurs, which markets are supported and whether a customer can reach a demo, change settings or delete an account.
“AI-powered” remains the operator’s wording until a concrete automated function and its controls can be observed.
The file notes authentication, recovery, withdrawal controls, API permissions, key handling, data sharing, incident records and the route for revoking access. A request for a seed phrase, private key or unnecessary withdrawal authority stops the normal test.
Minimum deposits, subscriptions, spreads, commissions, conversion charges, performance fees and withdrawal costs are taken from the plan and market actually visible during the check. We distinguish the payment recipient, asset custodian and execution venue instead of treating them as one party.
Deposit convenience does not stand in for a withdrawal test. Available methods, limits, stated timing, identity checks and any demand for another payment are recorded separately.
A return, win rate or backtest is usable only when its period, market, sample, benchmark, fees, slippage, drawdown and method can be identified. Screenshots, isolated trades and unexplained percentages remain claims, not representative results.
We look for simulated figures presented as live, favourable periods chosen after the fact and settings that cannot be reconstructed. A brief observation can reveal how controls behave, but not long-term profitability.
We identify the channel and entity responsible for account, payment and security issues. When a question is sent, the record includes its date, channel and answer. Without direct contact, the page makes no claim about response time.
User reports can point to a date, transaction, restriction or support exchange worth checking. Duplicate wording, coordinated bursts and unusual rating distributions affect the weight given to a pattern.
Repetition alone does not verify an allegation. Anonymous reports stay attributed, unnecessary personal data is excluded and factual claims are compared with records or reproducible product behaviour where possible.
A source is useful for the question it can answer. A company register can identify an entity but cannot prove product performance; an account screen can show a setting but not regulatory permission.
For each central point, the file keeps the URL or record reference, responsible entity, relevant jurisdiction, source date, access date and the narrow statement that the source supports. Volatile pages are captured when doing so is lawful and technically possible.
Relevant primary starting points include the ICANN registration data lookup, the European e-Justice business-register search and ESMA’s MiCA register. The relevant target-country authority and permission scope still take priority. An alert database is not exhaustive, and absence from it is not proof of authorisation.
Each published file names the access that was actually available. Reading public records is not presented as an account test, and a demo is not presented as a funded withdrawal.
| Level | What it covers | What it cannot establish |
|---|---|---|
| Desk check | Public records, legal documents, product pages, fee disclosures, support information and relevant third-party records. | Private account behaviour, live execution or whether an advertised workflow functions as described. |
| Public or demo walkthrough | An unauthenticated interface, accessible demo, onboarding preview or paper-trading environment. | Production custody, real funding, withdrawal behaviour or live-market execution. |
| Registered-account walkthrough | Registration, identity flow, account settings, permissions, support access and non-funded product controls where permitted. | Actual deposit handling, live trading quality, real fees or withdrawal completion. |
| Funded transaction trace | A disclosed deposit, configured strategy or order path and withdrawal attempt, with dates, market conditions, account mode and relevant settings recorded. | Future returns, safety under every condition or performance across a full market cycle. |
Geographic restrictions, identity requirements, unavailable software, account refusal, security concerns or disproportionate capital requirements may prevent a level from being completed. The limitation belongs in the assessment.
A confident tone cannot repair a weak source. The wording attached to a finding reflects what the file contains at the stated date and in the stated market.
Source strength and possible harm are recorded separately. A serious claim may remain open; a minor detail may be firmly confirmed.
Missing information, a contradiction and confirmed misconduct are different findings. Any of the following can end account testing or prevent a favourable conclusion:
Stopping a test protects the researcher and does not by itself prove criminal conduct. Published wording must match the evidence available.
Comparisons are meaningful only when the services were checked for the same market, period, customer type and level of access. Popularity, a promotional score or a long feature list does not supply that common basis.
The published conclusion records:
A favourable finding is limited to the named product, operator, jurisdiction, date and sources in that file. Services with different custody models or automation functions are not forced into one numerical ranking.
Dubaicat does not generate scores from a formula. A number appears only if the underlying observations create a visible, reproducible basis; identity, permission, custody, payment and withdrawal problems outweigh cosmetic strengths.
A correction names the affected statement, the date of the change, the reason and the replacement source. A change to ownership, legal status, access, prices, custody, withdrawals or the conclusion triggers a dated review rather than an unmarked rewrite.
An operator can send evidence or dispute a finding at contact@dubaicat.io. The reply is attributed and checked; it does not create a right to select the wording or remove a supported finding.
A supplied account, waived fee, sponsorship or affiliate relationship is disclosed beside the relevant material. A paid destination is not used to establish the official domain or operator, and payment cannot hide a risk or buy a favourable conclusion.
No. Company identity, permissions, warning lists and consumer protections are checked for the country and service named in the assessment. A result from one market is not carried into another without a separate record.
The assessment states where access ended and reports only the records or screens that were available. It does not describe a demo as a live account or infer a deposit, trade or withdrawal that did not occur.
A complaint can identify a transaction, restriction or support exchange to investigate. It changes a factual finding only when the underlying event can be supported; repetition by itself is not verification.
The page identifies the affected statement, the change date, the reason and the replacement source. An operator response is attributed and checked, but it does not create a right to remove a supported finding.