Verified, then logged.
Sun 22 Nov 2026 · 00:20 KST / 15:20 UTC한국어 | ENSearch /

Breaking04 / Policy

OECD.AI article lists what investors need to assess AI trust

The three elements and three policy areas are the authors' proposals, and no release date has been disclosed for a standard assessment tool for investors.

Two investors, seen from behind in a venture capital meeting room, study side-by-side comparison charts without text on a wall screen. AI-generated image

OECD.AI, the AI policy website of the Organisation for Economic Co-operation and Development, published a contributed article on 12 Oct on the conditions investors need to assess AI trustworthiness, citing a 61% share for AI companies in global venture investment in 2025. The article is the authors' opinion and not an OECD decision.

The article lists three things investors need: a practical framework, comparable data and credible verification. It divides the policy response into three areas: common signals, demand signals, and transparency and disclosure.

A notice attached to the article states that the authors' views do not necessarily match the official views of the OECD or its member countries. The three elements and three areas remain at the proposal stage.

OECD.AI's guidance defines the reporting framework of the Hiroshima AI Process, a G7 initiative on AI governance, as a common, voluntary structure through which organisations disclose their AI governance and risk management practices.

Korean providers. The AI Basic Act, Korea's framework statute on artificial intelligence, requires providers of high-impact and generative AI products and services to give advance notice, according to the version effective 21 Jul in the National Law Information Center, Korea's government statute database. That notice is owed to users, and disclosure to investors falls outside the provision.

The release date of a standard assessment tool for investors and the sample size of the venture-investor survey mentioned in the article have not been disclosed.