From 7d0e096d61859c819b5b660e64451699196f7453 Mon Sep 17 00:00:00 2001 From: artale Date: Tue, 7 Jul 2026 13:45:55 +0200 Subject: [PATCH] docs(tac): add antirez export-control corollary to insider-threat note MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Folds antirez's observation that LLM export controls are futile (weights leak, proxies bypass, enforcement requires dismantling encryption) into the existing insider-threat model note. Argues the real safeguard is agent-level sandboxing (OpenShell pattern), not model distribution control — assume any model can run on any hardware. --- plans/meta-prompts/loop_engineering.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/plans/meta-prompts/loop_engineering.md b/plans/meta-prompts/loop_engineering.md index 5dd70f5..b1d675a 100644 --- a/plans/meta-prompts/loop_engineering.md +++ b/plans/meta-prompts/loop_engineering.md @@ -143,6 +143,8 @@ Do not trust an agent because it is helpful, local, or model-authored. Trust onl Reference: Security Fest 2026, Dan Tentler, "Claude is your insider threat now" (YouTube `yvJYw2gR0cU`). +**Antirez corollary: model export controls are futile against the insider-threat model.** Model weights (few TB files) will inevitably leak; internet proxies render geographical blocks meaningless; enforcement would require dismantling encryption and global connectivity. The real question is not nationality-based restrictions but whether to distribute a powerful model to any general public at all — every large country has internal threats that could misuse leaked copies (antirez, July 2026). This strengthens the case for agent-level safeguards: if you cannot control which model an adversary runs, you must sandbox the agent, policy-gate every action with business impact, and assume any model could be running on any hardware. OpenShell's approach (secure agent runtime, secrets management outside sandbox, policy-based egress) is the correct architecture for this world. + ## External research anchors Ground agent plans against external production guidance, not just video notes: