Public Consultation on the Use of AI in the Financial Sector in Israel

By‎ Adv. Haim Ravia

Partner, Chair of the Cyber, Privacy & Copyright Group at Pearl Cohen Zedek Latzer Baratz

By‎ Adv. Dotan Hammer

Partner at the Cyber, Privacy & Copyright Group, Pearl Cohen Zedek Latzer Baratz

Reading time: 2 minutes
Save to "My Content"

An interdepartmental task force in Israel has released a draft interim report for public consultation addressing the integration of artificial intelligence (AI) in the financial sector, associated risks from predicted use in decision-making or human interactions, and proposed regulatory measures. The consultation period will end on December 15, 2024.

Key issues highlighted by the task force include transparency and explainability, human oversight, privacy protection, discrimination risks, and liability.

Transparency-related recommendations include informing consumers of the substantially impacting use of AI and possible alternatives unless the use of AI is obvious. These obligations will be integrated into existing notice obligations and subject to similar for and phrasing guidelines.

On explainability, the task force proposes to limit the obligation to provide explanations to special cases, such as when AI makes high-risk decisions or when human intervention is negligible. Such human intervention and oversight can be after the fact, except when the relevant decision substantially impacts an individual and technology is key in reaching that decision. In these cases, human intervention in the AI system should be in real-time.

The task force’s focus on privacy protection in AI led to multiple recommendations. AI implementers shall be instructed to assess the risks of identifiability of data subjects and execute data protection measures for anonymized data. The task force also recommended further strengthening the requirement for informed consent by adopting obligatory form and phrasing guidelines or requesting “layered consent”. The taskforce specifically requested the public’s opinion on whether inferred personal data should be subject to rights or obligations, such as the right of an individual to reasonable inferences, that are acceptable, relevant, and reliable.

The task force also addressed concerns over AI’s potential impact on discrimination, stating that while the laws prohibiting discrimination apply, the public’s opinion should be heard regarding implementing the disparate impact doctrine in U.S. law, which examines discrimination based on outcome rather than intent.

Finally, the task force recommended adopting the financial regulation approach to liability, where corporate organs and officers will be held liable in certain circumstances. Regulatory limitations on contractual waivers of liability were also recommended.

The report also underscores the importance of risk-based regulation due to the expected challenges of AI use in the financial sector, aligned with global standards like the EU's AI Act. The task force addressed additional actions that can be taken to increase the effectiveness of regulations, primarily government investment in innovation, and advancing regulatory certainty and oversight enterprises.