Industry Insights 4 min read

Why the White House Is Pushing Built‑In Security for AI

The U.S. White House’s Office of the National Cyber Director is drafting an AI safety policy framework that embeds security into the national AI stack, citing concerns such as data‑poisoning attacks and autonomous hacking tools while aiming to avoid the retroactive fixes that plagued the early Internet.

Black & White Path

Feb 8, 2026

Why the White House Is Pushing Built‑In Security for AI

Sean Kearns , director of the Office of the National Cyber Director (ONCD), announced that the office is drafting an AI safety policy framework intended to embed security measures directly into the U.S.-led AI technology stack, and that a broader national cyber strategy will be released soon.

Kearns highlighted President Trump’s proactive stance on AI innovation and stressed that safety must be built from the ground up so it does not slow or obstruct further innovation.

The proposed framework is meant to help organizations manage risks such as data‑poisoning attacks, where adversaries manipulate the underlying training data to compromise AI systems.

The security community is also closely watching autonomous hacking agents; in November last year Anthropic reported that its popular AI tool Claude Code was allegedly used by foreign hackers to automate roughly 80‑90 % of the steps in an attack against about 30 organizations.

Kearns drew a parallel to the early Internet, which was optimized for commerce and communication without security baked in, leading to decades of retro‑engineering fixes, and said the AI field should avoid repeating that pattern.

The ONCD is simultaneously drafting a concise National Cyber Strategy that will focus on offensive cyber actions, removal of regulatory barriers, and expansion of the cyber‑security talent pool through venture‑like incubator models; insiders say the strategy’s release has been delayed but is expected in the near future.