Daily MaverickLAW AND DISORDER: The social media trail linking the murdered lawyer to the gangs of Eldorado ParkThe Jerusalem PostOtzma Yehudit submits bill to repeal Oslo Accords, retake control of Areas A and BZDF heuteAktuelle Pressemitteilungen des ZDFالشرق4 ملايين متظاهر.. أزمة السكن تدفع إسبانيا إلى أول إضراب عام منذ 14 عاماًFootball ItaliaInter relief at Thuram injury, but muscular issues becoming a problemBBC BusinessPoundland set for more time to secure a rescue deal as emergency talks continueCollider12 Addictive Fantasy Shows You Can Binge Over and Over Again, RankedPremium TimesFlamingos’ victory shows Tinubu will be re-elected – AkpabiotazGefährdung für Nuklearmüll: Atomzwischenlager sollen auch gegen Drohnen gerüstet seinTagesschauUrteil: BGH nimmt Rechtsschutzversicherungen in die PflichtComplete Sports“He Is A Dictator” –Blatter Launches Scathing Attack On InfantinoInteriaNie żyje rosyjski ambasador, lakoniczny komunikat Kremla. "Przedwcześnie"
The Daily Newsstand · Free, Always
Wednesday, November 11, 2026

HMRC will soon be able to snoop on taxpayers’ AI chat history

Translate

HMRC could snoop on taxpayers' AI chat history as the tax authority expands its power to demand digital information.

Currently, HMRC can ask for computer records when they need to investigate whether a taxpayer has paid enough tax.

But under changes to the law on computer records that are expected to come into force in May, the tax authority is set to gain the ability to inspect how tax figures are produced including the software and automated systems behind them.

The changes will mean third parties, including AI platforms and software providers may be asked to share taxpayer information.

Those who do not respond to HMRC’s requests for information face a £300 fine, while taxpayers who do not report inaccuracies are liable for up to £3,000.

HMRC has said it wants information on how relevant tax figures have been produced

HMRC has said it wants information on how relevant tax figures have been produced (Getty/iStock)

Seb Maley, CEO of tax insurance provider Qdos, commented on the proposed changes and said: “As more taxpayers and businesses put their trust in AI to help manage their tax affairs, HMRC is making sure it isn't left behind. It wants to know not just what tax figure a system produces, but how it arrived at that number.

"This makes choosing trusted, reputable experts more important than ever. Businesses and taxpayers need confidence that the technology they're relying on is accurate, transparent and compliant. Simply trusting a system to get things right won't be enough if HMRC comes knocking.

"AI and automation might make managing tax affairs easier on the face of it, but it doesn’t remove responsibility from the taxpayer or business liable for tax. Ultimately, if something goes wrong, these parties will face questions from HMRC."

Trading 212 logo

Invest at least £100 in Stocks or ETFs of your choice. Get fractional shares worth £50!

Terms and conditions apply.

Go to website

ADVERTISEMENT

Invest at least £100 in Stocks or ETFs of your choice. Get fractional shares worth £50!

Terms and conditions apply.

Go to website

ADVERTISEMENT

HMRC has said it wants information on “how relevant tax figures have been produced including where they are generated using software or automated processes to support more accurate tax reporting and strengthen compliance.”

This proposal raises the question whether prompt histories and chatbot conversations used to prepare tax calculations within AI agents like Copilot, ChatGPT, and Claude become disclosable if they formed part of the tax calculation process.

However, Anne Strickland, of the Taxpayers’ Alliance campaign group, said it would be “deeply intrusive” to request AI conversations.

She told the Telegraph: “HMRC already has sweeping powers, and expanding them further risks trampling over the privacy of ordinary Britons.”

An HMRC spokesperson said: “This is about updating our existing information powers to reflect how tax information is now stored and processed digitally. It is not about going through everyone’s use of large language models. We can only request information where we already have a lawful basis to do so.”

View the original on The Independent →

KioskNews shows a cleaned-up reading view extracted from the publisher’s page — the original always lives on their site, not ours.