January 29, 2019

How does the theory of terrorism stack up against AML? Badly - finally a case in Kenya: Dusit

Finally, an actual financial system & terrorism case lands before the courts, relating to the Dusit attack. Is this a world first? I don't know because this conjunction is so rare, nobody's tracking it.

The essential gripe is that since 9/11 the financial world decided to slap the terrorism label on their compliance process. Yet to no avail. Very few cases, so small that they fall between bayesian cracks. So misdirected because terrorists have options, and they can adjust their approach to slip under they radar. Backfiring because the terrorists are already outside norms and will do as much damage as needed, thus further harming the financial system.

And so hopeless because your true terrorist doesn't care about being caught afterwards - he's either dead or sacrificed.

Anyway, that's the theory - anti-terrorism applied to the financial system simply won't work. Let's see how the theory stacks against the evidence.

A suspect linked to the Dusit terror attack received Sh9 million from South Africa in three months and sent it to Somalia, the Anti-Terror Police Unit have said. Twenty one people, including a GSU officer, were killed in the January 15 attack. The cash was received through M-Pesa.

So far so good. We have about $90,000 (100 Kenya shillings is 1 USD) sent through M-Pesa, a mobile money system in Kenya, allegedly related to the Dusit attack.

Hassan Abdi Nur has 52 M-Pesa agent accounts. Fourty seven were registered between October and December last year, each with a SIM card. He used different IDs to register the SIM cards.

So (1), the theory of terrorism predicts that the money will be moved safely, whatever the cost. We have a match. In order to move the money, 52 accounts were opened, at the cost of different IDs.

One curiosity here is the cost. In my long running series on the Cost of your Identity Theft we see (or I suggest) an average cost of an Identity set of around $1000. Which would amount to a cost of $52k for 50 odd sets. But this is high for a washed amount of $90k.

Either the terrorists don't care of the cost, or cost of dodgy ID is lower in Kenya, or the alleged middleman amortised the cost over other deals. Interesting for further investigation but not germane to this case.

Then (2), the theory of bayesian statistics and the "base rate fallacy" predict that no terrorists will ever be caught before the fact based on AML/KYC controls.

Clearly this is a match - the evidence is being compiled from after-the-fact forensics. Now, in this the Kenyan authorities are to be applauded for coming out and actually revealing what's going on. In the western world, there is too much of a tendency to hide behind "national secrets" and thus render outside scrutiny, the democratic imperative, an impossibility. One up for the Kenyans, let's keep this investigation transparent and before the courts.

Next (3). The theory predicts that follow the money is a useless tool.

Ambitham was in constant communication with slain lead attacker Ali Salim Gichunge, who died during the attack and his spouse Violent Kemunto Omwoyo.

[Inspector] Githaiga yesterday said Ambithamís phone led to his arrest on Tuesday after detectives established his communication with the Gichunges.

The police are following the social graph and arresting anyone involved. Having traced the phones, they then investigated the M-Pesa evidence, which provided many additional and interesting confirmatory facts.

Which is what they should do. But it was the contact information that cracked this case, not the financial flows. The contact information has always been available to them. And, where there is a credible case of terrorism as is in this case, the financial information has never been withheld. Again, the theory matches the evidence: follow the money is useless before the event, only confirmatory after the event.

Finally (4), the theory of unforeseen consequences says that the damage done by unintelligent responses will haunt the future of anti-terrorism efforts.

These are the agents that received the money, which was later withdrawn at the Diamond Trust Bank, Eastleigh branch, before it was wired to Somalia. ... The manager of the bank where Nur was withdrawing the money,, Sophia Mbogo, was arrested for failing to report Nurís suspicious transactions. Nur is said to have made huge withdrawals in short intervals, which Mbogo ought to have reported to relevant authorities, but there is no indication she did so.

Without wishing to compromise the investigation - this looks inept. Eastleigh is the Somali district of Nairobi. It's a bustling centre of trade. In some respects the Somalis are better traders than the Kenyans, and a lot of trade is done. And a lot of that is in cash, because the Kenyan banking system is ... not responsive. Lots of legitimate cash would move in and out of that bank branch.

Given the alleged fact that the money man had 52 M-Pesa accounts, he was certainly aware enough to run under the radar of the branch. Thresholds and actions by banks are no secret, especially by those motivated by terrorism to conduct any crime to find out - bribery, extortion, kidnapping are options.

Maybe there is evidence that the branch or the manager is "in" on the deal. Or maybe there is not, and the Kenyan police have just confirmed the theory that FATF anti-terrorism will do more damage. They've sent a message to all branches to drown their customers in pointless compliance, and to not cooperate with the police.

The Kenyan police had better get a clear and undeniable conviction against the branch manager, or they are going to rue the day. The next terrorist attack will surely be harder.

Posted by iang at January 29, 2019 01:33 AM
Post a comment

Remember personal info?

Hit preview to see your comment as it would be displayed.