Publications and working papers
19/06/21
- article
  • Astrid Bertrand
  • , Winston Maxwell
  • and Xavier Vamparys
Terrorist Financing: "Putting Artificial Intelligence And Data Sharing At The Center Of The Fight"
Astrid Bertrand, Winston Maxwell and Xavier Vamparys, researchers at Télécom Paris, explain in an article published in "Le Monde" that artificial intelligence can make anti-money laundering systems, which currently cost more than they earn in Europe, more effective.
11/06/21
- blog entry
  • Astrid Bertrand
The Us Anti-Money Laundering Act 2020 Supports The Deployment Of Technological Innovation And Machine Learning
The Anti-Money Laundering Act 2020 (AMLA) is a major watershed in the AML legislation of the US. It was passed by the Congress on January 1, 2021.
22/03/21
- blog entry
  • David Bounie
  • , Winston Maxwell
  • and Astrid Bertrand
Data Sharing Issues In The Finance Industry
The 3rd session of AI and finance with the ACPR and Télécom Paris tackled data sharing issues in the finance industry.
01/02/21
- blog entry
This Thing Called Fairness: The Importance Of An Interdisciplinary Approach
A paper by an interdisciplinary team of researchers at Berkeley summarizes why algorithmic biases necessarily require an interdisciplinary approach.
25/01/21
- blog entry
Interdisciplinary Perspectives On Algorithmic Biases In Finance
The January 11 edition of the ACPR and Télécom Paris "AI and Finance Mondays" focused on algorithmic biases. Second conference in a series dedicated to AI in finance, this webinar is part of the work of the Digital Finance and Explainable AI for Anti-money Laundering chair (XAI4AML).
27/10/20
- blog entry
  • Astrid Bertrand
Explainability Of Algorithms In The Fight Against Money Laundering
According to Europol's estimates, 200 billion euros of criminal funds circulate each year in Europe, and only 1% of these illicit financial flows are seized. Current anti-money laundering systems therefore have a great deal of room for improvement, particularly in order to counter increasingly sophisticated criminal methods. AI is proving to be a great help in improving the anti-money laundering system but, at the same time, it creates new forms of risk for individuals. This is where explainability comes in.
09/09/20
- blog entry
  • Astrid Bertrand
The Acpr’S Guidelines On Explainability: Clarifications And Ambiguities
The French banking regulator, the ACPR (Autorité de contrôle prudentiel et de résolution), recently published a report presenting guidelines on the governance of algorothms in the financial services sector. We analyze in this blog post these new requirements on AI, particularly aspects on explainability, reviewing the main clarifications brought by the report along with its limits.
06/08/20
- blog entry
  • Winston Maxwell
  • , Astrid Bertrand
  • and Xavier Vamparys
Current Anti-Money Laundering (Aml) Techniques Violate Fundamental Rights And Ai Would Make Things Worse
In a new paper, we have analyzed current AML systems as well as new AI techniques to determine whether they can satisfy the European fundamental rights principle of proportionality, a principle that has taken on new meaning as a result of the European Court of Justice’s Digital Rights Ireland and Tele2 Sverige – Watson cases. The question we addressed is whether proportionality requirements can be satisfied by AI-powered AML systems. To conduct our analysis we broke the proportionality test down into its various components. We then systematically applied each of the steps of the proportionality test to AML systems, both current rule-based AML systems and then to AI-enhanced systems. Our findings are that current AML systems fail the proportionality test in five respects. AI makes the failures more acute, but does not fundamentally change the reasons for the underlying problems. The one area where AI adds a new specific problem compared to current systems is algorithmic explainability. Our paper has been selected for presentation at the July 17 ICML2020 Law and Machine Learning workshop.
06/08/20
- Conference paper
  • Winston Maxwell
  • , Astrid Bertrand
  • and Xavier Vamparys
Are Ai-Based Anti-Money Laundering Systems Compatible With Fundamental Rights?
In a new paper presented at the July 17 ICML2020 Law and Machine Learning workshop, we have analyzed current AML systems as well as new AI techniques to determine whether they can satisfy the European fundamental rights principle of proportionality. Here is the abstract: Anti-money laundering and countering the financing of terrorism (AML) laws require banks to deploy transaction monitoring systems (TMSs) to detect suspicious activity of bank customers and report the activity to law enforcement authorities. Because the monitoring of customer data to detect money laundering interferes with fundamental rights, AML systems must comply with the proportionality test under European fundamental rights law, as most recently expressed by the Court of Justice of the European Union (CJEU) in the Digital Rights Ireland and Tele2 Sverige - Watson cases. To our knowledge there has been no analysis as to whether AML systems are compliant with the proportionality test as expressed in these latest cases. Understanding how the proportionality test applies to current AML systems is all the more important as banks and regulators consider moving to AI-based tools to detect suspicious transactions. The objective of this paper is twofold: to study whether current AML systems are compliant with the proportionality test, and to study whether a move towards AI in AML systems could exacerbate the proportionality problems. Where possible, we suggest cures to the proportionality problems identified.
02/04/20
- blog entry
  • Winston Maxwell
  • and Xavier Vamparys
Netherlands Welfare Case Sheds Light On Explainable Ai For Aml-Cft
The District Court of The Hague, Netherlands found that the government’s use of artificial intelligence (AI) to identify welfare fraud violated European human rights because the system lacked sufficient transparency and explainability. 1 As we discuss below, the court applied the EU principle of proportionality to the anti-fraud system and found the system lacking in adequate human rights safeguards. Anti-money laundering/countering the financing of terrorism (AML-CFT) measures must also satisfy the EU principle of proportionality. The Hague court’s reasoning in the welfare fraud case suggests that the use of opaque algorithms in AML-CFT systems could compromise their legality under human rights principles as well as under Europe’s General Data Protection Regulation (GDPR).
02/09/19
- article
  • David Bounie
  • and Winston Maxwell
Is Explainability Of Algorithms A Fundamental Right?
“The demand for transparency on the functioning of algorithms must be addressed with discernment”, assert, in a column to “Le Monde”, researchers David Bounie and Winston Maxwell.