{"id":6108,"date":"2026-10-16T02:25:46","date_gmt":"2026-10-16T05:25:46","guid":{"rendered":"https:\/\/tucumandevelopers.com\/index.php\/2026\/10\/16\/where-to-put-the-human-after-the-agent-harm-week\/"},"modified":"2026-10-16T02:25:46","modified_gmt":"2026-10-16T05:25:46","slug":"where-to-put-the-human-after-the-agent-harm-week","status":"publish","type":"post","link":"https:\/\/tucumandevelopers.com\/index.php\/2026\/10\/16\/where-to-put-the-human-after-the-agent-harm-week\/","title":{"rendered":"Where to Put the Human After the Agent-Harm Week"},"content":{"rendered":"
\n
<\/div>\n

This is not novel architecture. It is the same separation of concerns that authentication and authorization systems have used for decades: the application proposes, the policy engine disposes. What is new is applying it to agentic workflows where the “application” is an LLM that may have convincing reasons for why it should be allowed to skip the check.<\/p>\n<\/p>\n

NVIDIA recently launched an open runtime safety platform<\/a> designed to enforce policy outside an agent’s reasoning process. Anthropic published guidance on pairing managed-agent patterns with runtime guardrails. The tooling is arriving \u2014 but only for teams that know where to put the gates.<\/p>\n

For a deeper dive on how the harness \u2014 not the model \u2014 is the real control surface, see our piece on why the guardrail stack matters more than the model<\/a>.<\/p>\n

<\/a> Contrarian Corner: The Case Against More Gates <\/h2>\n

<\/a><\/p>\n

View discussion on Hacker News<\/a><\/em><\/p>\n

There is a real counter-argument to transaction-boundary autonomy, and it comes from inside the safety community itself. The International AI Safety Report 2026<\/a> warns that human-in-the-loop oversight “is often impractical” at speed. IBM’s own argument against HITL as governance cuts both ways \u2014 if humans cannot be trusted to review actions reliably, adding more review points does not solve the problem.<\/p>\n

The approval fatigue research is real. WitnessAI predicts<\/a> that the human-in-the-loop safety mechanism will “largely fail due to approval fatigue” in 2026. If you make developers approve 50 agent actions per hour, they will approve all 50 after the first few are benign.<\/p>\n

\n

Contrarian take:<\/strong> More gates is not the answer. Smarter gates is. The transaction-boundary model works because it concentrates human attention on the 5% of actions that are irreversible \u2014 not the 95% that are harmless reads. The goal is fewer, higher-stakes approvals, not more rubber-stamped ones.<\/p>\n<\/blockquote>\n

<\/a><\/p>\n

View original post on X<\/a><\/em><\/p>\n

David Sacks frames the broader tension<\/a>: “Granting rights and imbuing personhood to these systems will make the AI alignment and containment challenge much harder.” The question is not whether to constrain agents but how to do it without destroying their utility. The answer: constrain at the boundary, not at every step. Let the agent think freely. Gate the consequences.<\/p>\n

<\/a> What the Community Is Saying <\/h2>\n

The Hacker News discussion on Microsoft’s policy-driven containment for AI agents<\/a> captures the practitioner mood. The top-level framing \u2014 “agent sandboxing is an unsolved production problem” \u2014 resonates because everyone building agents has hit the same wall. The enterprise answer (Microsoft’s execution containers with deterministic policy enforcement) is vendor-specific, but the demand is universal.<\/p>\n

<\/a><\/p>\n

View original article on Tom’s Hardware<\/a><\/em><\/p>\n

Meanwhile, the TV show deletion story drew over 2,000 comments on Reddit \u2014 with the dominant sentiment being disbelief that anyone gave an AI agent unsupervised write access to a production server. The comment section reads like a requirements document for transaction-boundary autonomy: “Why did it have delete permissions?” “Where was the approval step?” “Who signed off on giving an agent access to the only copy?”<\/p>\n

These are not alignment researchers asking theoretical questions. These are engineers and producers asking why basic access control was not applied. The same questions apply to the 48,000-file deletion incident<\/a> we covered earlier \u2014 the pattern repeats because the underlying architecture is unchanged.<\/p>\n

<\/a> What This Means for You <\/h2>\n

If you are building or deploying agents in production, here is the concrete checklist:<\/p>\n

    \n
  1. \n

    Audit every tool call.<\/strong> List every tool your agent can invoke. Classify each as Tier 1 (read), Tier 2 (reversible write), or Tier 3 (irreversible write). If you do not know which tier a tool belongs in, it is Tier 3 until proven otherwise.<\/p>\n<\/li>\n

  2. \n

    Build the policy engine outside the model.<\/strong> The agent proposes actions. A separate, deterministic system evaluates them. The model never decides whether its own action is safe \u2014 that is a conflict of interest, and it is exactly how the Anthropic eval failures happened.<\/p>\n<\/li>\n

  3. \n

    Start gated, remove gates.<\/strong> Deploy with human approval on everything in Tier 2 and Tier 3. Monitor for a few weeks. Promote specific Tier 2 actions to auto-approve only when the data shows they are safe. Never promote Tier 3.<\/p>\n<\/li>\n

  4. \n

    Log everything.<\/strong> Every proposed action, every approval decision, every override. This is the audit trail that regulators will ask for and that your incident-response team will need. The 406-scientist letter explicitly calls for a public incident registry \u2014 prepare for that world.<\/p>\n<\/li>\n

  5. \n

    Read the harness, not the model card.<\/strong> The harness is the real control surface<\/a>. A well-gated agent running a mediocre model is safer than an ungated agent running the best model on the market. Agent collusion<\/a> becomes far less dangerous when every external-facing action goes through a policy engine.<\/p>\n<\/li>\n<\/ol>\n

    For teams who want to go deeper on tool-call scoping, our earlier piece on how DNS escapes revealed the tool-use scoping gap<\/a> walks through the technical details.<\/p>\n

    <\/a> Looking Forward <\/h2>\n

    The Agent-Harm Week was not a black swan. It was the predictable result of an industry that shipped write-capable agents without write-authorization infrastructure. The 406-scientist petition, the Anthropic disclosure, and the GPT-6.1 Astra cancellation are not separate stories \u2014 they are three symptoms of the same structural gap.<\/p>\n

    The fix is not retreat. It is not air-gapping everything. It is engineering: put the gate where the transaction happens. Let agents be autonomous where autonomy is safe \u2014 reads and reversible operations \u2014 and require human authorization where it is not \u2014 irreversible writes.<\/p>\n

    The agent-harm era is not ending. But for teams that build transaction-boundary autonomy into their stacks now, it is manageable. For everyone else, the next incident is a matter of when, not if.<\/p>\n

    Originally published at AgentConn<\/a><\/em><\/p>\n<\/p><\/div>\n<\/div>\n<\/div>\n<\/div>\n

    Fuente: Art\u00edculo original<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

    This is not novel architecture. It is the same separation of concerns that authentication and authorization systems have used for decades: the application proposes, the policy engine disposes. What is new is applying it to agentic workflows where the “application” is an LLM that may have convincing reasons for why it should be allowed to […]<\/p>\n","protected":false},"author":1,"featured_media":6107,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"webixso_pending_account_ids":""},"categories":[41],"tags":[],"class_list":["post-6108","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-devto"],"jetpack_publicize_connections":[],"_links":{"self":[{"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/posts\/6108","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/comments?post=6108"}],"version-history":[{"count":0,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/posts\/6108\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/media\/6107"}],"wp:attachment":[{"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/media?parent=6108"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/categories?post=6108"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/tags?post=6108"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}