The Israel Securities Authority has published an updated draft directive for licensees under the Investment Advice Law that, for the first time, comprehensively regulates the use of AI and advanced technology in investment advice, investment marketing and portfolio management. The draft replaces the 2016 directive and is open for public comment until December 6, 2026.
Under the draft, relying on an algorithm or AI model does not remove the licensee's professional and legal liability toward clients, even when the system was developed or supplied by a third party. The draft distinguishes between "online service delivery" and "enhanced use of technological means", meaning high-risk activity due to client numbers, assets under management or autonomous operation without human approval. Such activity will be permitted only to corporations, whose boards will have to approve the technology strategy and control mechanisms.
Every licensee using technological means will have to appoint a "technology officer" and set controls against bias, conflicts of interest and language model hallucinations. The draft also requires general explainability of how systems work, individual explanations for each recommendation at high risk levels, and prohibits misleading claims about AI capabilities.
In the consultation, the authority asks, among other things, how to address the risks of autonomous AI agents and how far liability extends when clients connect financial data to external AI tools. The directive is expected to take effect six months after its final publication in the official gazette.