AI vendor due diligence

Vanaxity insights tagged AI vendor due diligence - notes on AI content, SEO, GEO and AEO from Van Data Team.

3D illustration of a glass wall with a lit grid and an open arch, a stream of light passing through the arch while crates edged in amber wait at the wall.
China AI Export Rules
Governance & ROIChina AI export controlsAI export controls
Governance & ROI · Oct 19, 2026

China AI Export Controls: What Marketing Teams Should Check

On October 16, 2026, China's commerce and science ministries proposed adding 40 technologies to the country's export control list, 23 of them in AI. Core technology for models with at least 2 trillion parameters, 10^26 training FLOPs or a global top-10 ranking would need a license to leave China, including weights and source code. A registration step for these exports is already in force, and comments on the full list close on November 15. Crucially for marketers, already-public open-source models and use through APIs, apps and SaaS tools are exempt. Most marketing teams won't notice a change, but those whose AI vendors license private Chinese models or deploy them overseas should ask questions now.

3D isometric tech illustration of an architectural pedestal with a legislative parchment plaque and three concentric compliance rings.
A Pledge, Not a Law
Governance & ROIWhite House AI safety pledgeAI safety pledge
Governance & ROI · Sep 30, 2026

White House AI Safety Pledge: What It Means for Marketers

On September 29, 2026, President Trump and leaders of Anthropic, OpenAI, Google, Meta, xAI and Nvidia signed a voluntary accord, the Joint Commitment on Frontier Responsibilities. The companies promised robust internal controls, internal teams to check those controls, independent external auditors, board oversight and regular meetings on safety standards. The accord isn't legally binding, names no auditors and sets no public reporting rules, and critics called it vague. For marketers, the direct impact is small, but it raises the bar for vendor due diligence, makes safety claims riskier, and adds a new term, super intelligence, that federal agencies must now use.