The Financial Conduct Authority (FCA) has published its first public report on Horizon Scan technology, a forward-looking document grouping emerging risks under three headings: personal intelligence, artificial crime and programmable finance.
The organizer was careful about what the document was. the Financial Supervision Authority It is “not a set of expectations or regulatory guidance,” he said, but a map of plausible ways in which technologies can combine to benefit consumers, businesses and markets by 2030.
Read the full FM Intelligence analysis on the DataLab portal: Three AI transformations and a $40 billion fraud forecast: Read the FCA’s first horizon survey..
Synthetic crime could push AI fraud to $40 billion
The report describes that AI lowers the barrier to large-scale fraud and makes detecting fabricated evidence more difficult. It marks the move from manipulating what people see and hear to the end Deep fake audio and videoto manipulate how they judge what is right.
Deloitte Financial Services estimates that AI fraud losses in the United States could rise from $12.3 billion in 2023 to $40 billion by 2027.
Fraud losses in the UK exceeded £1.17 billion in 2024, according to UK Finance, with authorized payment fraud at £450.7 million.
The detection problem is already visible. Account takeover scams linked to artificial intelligence Increased by 250% in 2024and regulators including the FSCA in South Africa AI voice cloning undermines voiceprint authentication, warned.
When the next agent is an AI agent
The second theme, Personal Intelligence, describes AI agents who have become the primary interface between consumers and businesses. The Financial Conduct Authority delineates an “agent economy” in which an agent researches, compares and transacts for the client.
It outlines three steps: utilities that compare and pre-populate, advisory agents that recommend, and a “do it for me” mode that works within established boundaries.
This is not hypothetical. eToro Now It allows investors to delegate trades to artificial intelligence agents Within a specific budget, MasterCard and Santander have done this Trigger a live payment executed by the AI agent Within an organized framework.
An open question about compliance is who consents when software runs on someone’s behalf.
Token assets reach $18.6 billion with plumbing changes
The third topic, Programmable Finance, covers tokenization, stablecoins and smart contracts moving from pilots to national strategies.
Real tokenized assets on-chain have risen to around $18.6 billion through 2025, according to industry trackers, and the FCA defines the destination as “TradFi with protocol capabilities” rather than a full shift to decentralized finance.
The trend is already starting to move in the UK. Its Financial Supervision Authority It selected four companies to conduct experiments on stablecoins
Ahead of its 2027 crypto system, a business that sits alongside the Digital Securities Sandbox it runs with the Bank of England.
Discovery turns towards the discovery of suspicious perfection
For compliance and financial crimes teams, the direct line is detection. The report notes that the fraud signal may move from clearly wrong to what it calls “suspicious completeness,” which complicates controls designed to report anomalies.
The examination also points out that there is a risk of concentration, as many companies rely on the same AI platforms and cloud providers, so one common bug could hit several at once.
A full breakdown of FM Intelligence, including the data behind each theme, scenario ranges, and forecast limits, is on the DataLab portal: Read the analysis here.
This article was written by Damian Schmil at www.financemagnates.com.
Source link
