AI Risk Wire

Cases › other › LR-2026-0671

Musk and Rupert deepfake ads drew victims into $61 million investment scheme

Preliminary Why: Publisher not yet named · subject to change

Developing: a major or severe case captured in the last 72 hours. Later changes are logged on the corrections page.

Status: no action reported

Severity major
Reported 21 Aug 2026
Industry financial services

No source states: loss or records, developer and model, agent tool, could do, deployer.

Severity major: tier, because a regulator is involved: the FSCA investigated and fined Banxso and referred its findings to the Hawks.

South Africa's Financial Sector Conduct Authority (FSCA) found that Banxso, an online trading platform, used deepfake advertisements showing Johann Rupert and Elon Musk to promote an offering called Immediate Matrix. Individuals who responded were redirected to Banxso representatives and encouraged to trade mainly contracts for difference. The FSCA found that client funds were not placed with legitimate liquidity providers, were commingled, and were partly misappropriated. Banxso denies creating the advertisements and says it was itself a victim of hacking. The FSCA fined Banxso and individuals, withdrew Banxso's licence and referred its findings to the Hawks. In March 2026 the Western Cape High Court placed Banxso in final liquidation. The Financial Services Tribunal dismissed licence reconsideration applications in December 2025, and other challenges remained pending.

What happened

1 · What it reached or didThe losses eventually linked to the operation reached approximately $61 million, according to the FSCA.
2 · How it was caughtThe FSCA launched an investigation in March 2024 after receiving information about the deepfake campaign.

What was done

No source states an action or fix.

Risks

Read from the sources; not yet reviewed by a person.

Controls

Not yet read for controls. Controls that mitigate the primary risk:

Model

Marking machine-made output · Abuse monitoring and account enforcement · Content safety classifiers

Application

People told they are dealing with AI

Infrastructure / target

Phishing and fraud filters

Evidence

No source states: what the AI was doing, what went wrong, when it happened, how many records were exposed, a count of the other parties affected, any court case, the cause, what set it off, what was done.

Show the full account: 6 items from the primary sources

What happened, in order

  1. 2024-03 FSCA launches investigation after receiving information about the deepfake campaign.
  2. 2025-09 Banxso and its key individuals apply to the Financial Services Tribunal for reconsideration of the licence withdrawal.
  3. 2025-12 The Financial Services Tribunal dismisses the reconsideration applications.
  4. 2026-02 Five Banxso representatives lodge fresh applications for reconsideration of the debarments and administrative penalties.
  5. 2026-03 The Western Cape High Court places Banxso under final liquidation.
  6. July 31 The FSCA's 2025/26 Regulatory Actions Report is published.
Show every field, with its source and grade

What happened

Earliest source we hold
2026-08-21
D inferred by us from the described behaviour, reason shown
What went wrong
other
D inferred by us from the described behaviour, reason shown
primary risk
RK19
D inferred by us from the described behaviour, reason shown

Impact

Severity
major
D inferred by us from the described behaviour, reason shown
Why this severity
tier, because a regulator is involved: the FSCA investigated and fined Banxso and referred its findings to the Hawks.
D inferred by us from the described behaviour, reason shown

Context

Industry
financial services
C one secondary source states it
Country
South Africa
C one secondary source states it

Notes on gaps

The sources do not name the AI tool or model that produced the deepfakes, so system fields are left out. The date the incident occurred is not stated. The number of affected investors is not stated.

LR-2026-0671 · version 1 · published 2026-11-01 after automatic checks; not yet checked field by field by a person · JSON