Blog

Privacy, your data, and what AI keeps.

Practical insights navigating AI in regulated environments.

Switching from ChatGPT to Hush: what changes and what doesn't

Your prompts work the same way. Your data goes somewhere fundamentally different. A practical migration guide for the web app and the API.

Read →

Private audio transcription for clinical notes

Upload a recording, get a transcript in seconds, and no audio is retained on any server. GPU-accelerated speech-to-text on hardware we own, not a US cloud.

Read →

How a solicitor reviews a contract with Hush AI

Upload or paste a contract, get clause extraction and risk flags in seconds. No client data stored, no privilege waived. A step-by-step walkthrough.

Read →

How a GP drafts a referral letter with Hush AI

Paste the clinical context, review the draft, copy it back. Five steps, under a minute, and no patient data stored on any server.

Read →

Washington switched off the world's most powerful public AI. For everyone.

A US directive disabled Claude Fable 5 and Mythos 5 for every foreign national, then every customer worldwide. The vendor objected and complied anyway. Why a foreign order can't switch Hush off.

Read →

DSPT and AI: what GP practices need to know

The Data Security and Protection Toolkit has no box marked "AI", which is exactly why staff use of it can quietly make your answers inaccurate. Six things that keep AI defensible in your submission.

Read →

An AI-use policy for law firms (and why "don't" isn't one)

Fee-earners are already using AI. A short, clear firm policy, approved tools, a confidentiality rule, mandatory review, protects privilege better than a ban nobody follows. What to include.

Read →

AI for researchers: keeping unpublished data unpublished

Pasting interview transcripts or unpublished results into a chatbot is a data-governance decision your ethics approval may not cover. Three checks before AI touches research data, plus the integrity point researchers can't skip.

Read →

AI and client confidentiality: what accountants and advisers must check

Before AI touches client financial data, check three things: where it's processed, who can be compelled to hand it over, and whether it trains on your inputs. A short, practical guide for firms.

Read →

Your HR team is pasting employee data into AI. Here's the governance gap.

Grievances, sickness notes and performance concerns dropped into a chatbot are special-category data in an unassessed tool. Why HR is the sharp end of shadow AI, and what good looks like.

Read →

Seven questions to ask any AI supplier (including us)

Before you let an AI tool near client, patient or employee data, put these seven questions to its supplier in writing, and know what a good answer sounds like. Works on us too.

Read →

What actually goes in a practice AI policy (and why silence isn't one)

A practice AI policy doesn't need to be long, it needs to be clear. The seven things to cover, in plain language, with a free template you can adapt in an afternoon.

Read →

A quarter of GPs use generative AI at work. Here's what your practice should do about it.

One in four UK GPs already uses generative AI in clinical work, 95% with no training, almost none with a practice policy behind them. Three practical steps for your practice this month, with a free policy template and DPIA.

Read →

Microsoft told the French Senate it cannot guarantee European data stays out of US hands

Under oath, Microsoft France's legal director confirmed what the CLOUD Act already said: residency is geography, jurisdiction is the law. What that means for UK practices and law firms choosing AI tools, and the one question to ask any vendor.

Read →

Why Hush is not an ambient scribe (and what it does instead)

Where the MHRA medical-device boundary sits, why Hush deliberately stays on the administrative side of it, what we give up by doing so, and why, if you want a scribe, you should buy a scribe.

Read →

Your Fee Earners Used ChatGPT Today. Client Privilege May Already Be Gone.

The Upper Tribunal observed, in terms, in Munir v SSHD [2026] UKUT 00081 (IAC) that using public AI tools with client data can waive legal professional privilege. 59% of UK fee earners use unapproved AI on client data.

Read →

Is Your AI Provider Subject to the US CLOUD Act?

Most UK organisations believe their data is protected because it is "hosted in the UK." This belief is incorrect. The CLOUD Act applies to companies, not data centres, and the consequences for regulated professionals are significant.

Read →