Privacy

Last updated: 29 September 2026.

Contact admin@abcintel.uk about privacy, corrections or human review of a report. No account or complaint form is required.

Reader experience

The service does not offer reader accounts, subscriptions, comments or payment facilities. It does not use advertising profiles, cross-site tracking, fingerprinting, precise location or identified reading histories.

Aggregate usage statistics

First-party statistical counters help improve the service. They count public page views and article openings, approximate jurisdiction where the hosting edge supplies it, broad device categories and referral categories. They do not measure unique visitors or sessions. Search terms, URL query strings, IP addresses, raw browser identifiers, full referrer URLs and individual event records are not stored in the analytics database. Separate daily totals cannot be joined into a reader journey.

These counters run within the existing hosting platform; no third-party analytics script or advertising service is added. Buckets older than 90 days are removed when collection next runs; this is not an independent timed-deletion job. Network information is necessarily processed by the hosting infrastructure to serve requests. Separate request/security logs are handled by the hosting provider and are not part of these statistical counters.

Your analytics choice

You can stop these statistical counters at any time using the analytics preference on the Cookies page. We also honour Do Not Track and Global Privacy Control signals. A preference cookie is set only when you ask to opt out; it contains no identifier. This choice does not disable necessary security logging.

Editorial information and automated publication

We report selected bribery, corruption and related integrity-case developments from public official authorities, courts and regulators. Reports and supporting records may contain names, professional roles, organisations, jurisdictions, alleged conduct, procedural status, findings and outcomes. The service does not produce personal background assessments. Source attribution, minimisation and the distinction between allegations and established findings are essential.

AI assists selection, extraction, drafting and validation. Qualifying reports can be published automatically after evidence and publication checks; not every report is checked by a person. You can challenge an article, supply evidence or request human review by email. Publication can affect reputation and can reach readers worldwide and search engines. We do not claim exhaustive monitoring of later outcomes.

Editorial basis and criminal-offence information

We use Article 6(1)(f) UK GDPR for our legitimate interests in accurate public-interest journalism, explaining enforcement developments and operating the publication. For qualifying reporting about identifiable people’s alleged or established offences, we rely on Data Protection Act 2018 Schedule 1 paragraph 13, read with paragraph 36, where its conditions are met. Article 6 legitimate interests alone does not authorise criminal-offence processing. Paragraph 13 does not require an Appropriate Policy Document; this is not a blanket exemption from UK GDPR or an authorisation for unrelated uses.

Our policy requires a reasonable public-interest belief and case-specific consideration of accuracy, necessity, naming and rights; official publication by another body is not a substitute for those decisions. Any journalism exemption or refusal of rights must be justified individually.

Administration and correspondence

Private administration requires authentication. Authorised editor identities and actions are audited. Correspondence is used to handle enquiries, corrections, complaints and rights requests, with legitimate interests in responding and maintaining security and the applicable legal obligation where handling is required by law. Do not send unnecessary identity documents, privileged or confidential material.

Providers and information recipients

The service uses OpenAI Sites / Cloudflare for hosting and aggregate storage, Supabase for editorial records, and OpenAI for server-side editorial assistance. The contact mailbox uses Apple iCloud Mail. The director monitors it, with absence cover by the company secretary. The Supabase project’s configured database region is London, but that does not establish UK-only processing by all providers. Sites does not provide data-residency guarantees.

Supabase holds original source/provenance records, permitted extracts, assessments, article drafts, published articles and editorial audit history. Server-side OpenAI requests contain selected source material and editorial instructions; public-reader analytics are not included. The OpenAI account’s voluntary sharing of API inputs, outputs, feedback and evaluation data for model improvement is disabled. Requests disable response storage, which is not a claim that all provider safety logs are absent: OpenAI describes default abuse-monitoring retention of up to 30 days, with exceptions. Hosting request logs can contain network addresses, request paths, headers and browser information independently of aggregate analytics.

Retention and international arrangements

We retain reports and necessary evidence while their journalistic or historical purpose justifies it, taking account of accuracy, later outcomes and people’s rights. Our adopted schedule calls for review of unpublished material at 90 days and normal removal or minimisation by 180 days; routine correspondence 12 months after closure; and minimal complaint/decision records three years after closure with annual review. A recorded continuing editorial need or legal dispute may justify longer retention, with review dates.

These periods are implemented through a review process, not automatic editorial expiry. Aggregate analytics retains its separate 90-day collection-triggered cleanup. Supabase confirms that its Data Processing Addendum forms part of the account terms; it includes Standard Contractual Clauses and a UK Addendum for relevant international transfers. OpenAI’s published Sites and API data-processing terms also provide SCCs with the UK Addendum for covered UK data. These arrangements do not promise UK-only storage.

Copies are available in Supabase’s DPA, the Sites DPA and OpenAI’s API DPA, or by contacting admin@abcintel.uk. Apple’s privacy policy describes its international processing and Standard Contractual Clauses and explains how to obtain a copy. Provider security logs and retained copies follow the applicable provider terms and legal obligations, separately from our own review timetable. We do not promise a fixed deletion period for every provider log or copy.

Your rights and complaints

You may request access, correction, deletion or restriction of personal information, or object to its use. You may request human consideration of disputed automated publication. Rights depend on the circumstances, including freedom of expression and any specifically justified journalism exemption; requests require individual consideration and an explained decision.

Rights requests normally require a response within one calendar month, subject to lawful adjustments. Data-protection complaints must be acknowledged within 30 days and addressed without undue delay. We aim to acknowledge correspondence within two working days. You may also complain to the Information Commissioner’s Office or seek a court remedy. A public privacy page does not itself resolve any individual-notice obligations.