pith:SSWLMT43
The Two Boundaries: Why Behavioral AI Governance Fails Structurally
AI systems governing effects must make their capability boundary identical to the governance boundary or else risk and theater are inevitable.
arxiv:2604.27292 v3 · 2026-04-30 · cs.AI
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{SSWLMT43QJITZXPML7UNPCWYSI}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
Rice's theorem (1953) proves the gap is undecidable in the general case for any Turing-complete architecture that attempts to govern effects behaviorally: no algorithm can decide non-trivial semantic properties of arbitrary programs, including the property 'this program's effects comply with the governance policy.' We propose coterminous governance as the testable criterion for any AI governance system: either the two boundaries are provably identical, or risk and theater are structurally inevitable.
That deployed AI systems attempting behavioral governance of effects can be accurately modeled as arbitrary Turing-complete programs whose semantic properties must be decided post-hoc, and that separating computation from effect is both feasible and sufficient to achieve coterminous boundaries in practice.
Behavioral governance of AI effects is undecidable for Turing-complete architectures, making coterminous boundaries via computation-effect separation the only structural solution rather than post-hoc layers.
Receipt and verification
| First computed | 2026-05-27T01:05:55.280121Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
94acb64f9b82513cddec5fe8d78ad89220a81949cd11307baa9d9682dfc00904
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/SSWLMT43QJITZXPML7UNPCWYSI \
| jq -c '.canonical_record' \
| python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 94acb64f9b82513cddec5fe8d78ad89220a81949cd11307baa9d9682dfc00904
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "fe36b174680d6d3993a92d5ea1205eabb081b1413ff60e835e3b43a679c155f9",
"cross_cats_sorted": [],
"license": "http://creativecommons.org/licenses/by/4.0/",
"primary_cat": "cs.AI",
"submitted_at": "2026-04-30T01:12:32Z",
"title_canon_sha256": "30ac9e817b2737bcab548dc167b73f2d4802ca394d4c2b780bad83abc3811bb8"
},
"schema_version": "1.0",
"source": {
"id": "2604.27292",
"kind": "arxiv",
"version": 3
}
}