Scope of the Generalized Second Law

The generalized second law (GSL) says that horizon area and exterior quantum entropy combine into a nondecreasing generalized entropy. At the 2026-08-20 evidence cutoff, this is an established theorem for important semiclassical causal- and Killing-horizon domains, not a universal theorem for every trapping horizon, quantum state, or nonperturbative theory of gravity. The productive frontier is therefore the boundary of the proof, not the slogan that “total entropy increases.”

Helpful background. Proof Strategies for the Generalized Second Law develops relative-entropy monotonicity, the horizon-area relation, and the conditional theorem used here.

Causal horizons and the finite proposition

For a cut $C$ of a causal horizon, the renormalized quantity is schematically

$$S_{\mathrm{gen}}[C] =\frac{A[C]}{4G_{\mathrm{ren}}\hbar} +S_{\mathrm{out,ren}}[C] +S_{\mathrm{local}}[C].$$

The finite question is whether $S_{\mathrm{gen}}[C_2]\ge S_{\mathrm{gen}}[C_1]$ when $C_2$ lies to the future of $C_1$, for the stated horizon, field algebra, state class, renormalization prescription, and perturbative order. These qualifiers do real work. An event or causal horizon supplies nested exterior regions; an apparent or trapping horizon generally does not supply the same algebra inclusion. A semiclassical state with controlled backreaction is not an arbitrary state of full quantum gravity.

Assessment at 2026-08-20	Meaning
Question status	Partially resolved: broad semiclassical causal/Killing-horizon cells are proved; quasi-local and fully quantum versions remain open
Evidence status	Established in the named theorem domains
Confidence	High, because the positive statements are proofs and the principal limitations are stated by their authors
Domain	Semiclassical QFT on causal or Killing horizons with the state, algebra, asymptotics, and expansion assumptions stated below
What could change the conclusion	A proof/counterexample beyond leading semiclassical order or for a broader horizon class

A starting project: trace the sign in one GSL proof

Choose this direction if you can work with density operators, relative entropy, null congruences, and the difference between an operator algebra and a tensor factor. Begin with the linked proof-strategy background. Then read Faulkner and Speranza 2024, §§3.1–3.3, especially Eqs. (3.57)–(3.58) before its correction analysis in §5. Extract the nested algebras, reference state, cut-independent terms, and gravitational constraint used to translate an information inequality into an entropy law.

The first project is an annotated derivation of the entropy difference for two cuts in that semiclassical setting. Reconstruct each equality from its assumptions and determine the direction of the algebra inclusion without copying the final sign. Check the stationary limit and identify where the same derivation would fail for an arbitrary apparent horizon. A finite-dimensional partial-trace example can help diagnose the information-theoretic sign, but it cannot verify the gravitational theorem or the continuum algebra construction.

Deliver a short proof reconstruction with an explicit hypothesis-to-step map and one worked limiting case. This is primarily an analytic project; symbolic algebra is optional, while access to a supervisor familiar with algebraic QFT is much more valuable than additional computing power. If modular theory is still unfamiliar, stop at the relative-entropy and geometric-constraint bridge and use that as a reading project.

A candidate research extension is to isolate one correction term or one altered state class for which a sign estimate might be possible. Check the later literature before deciding that estimate is open. If the proposed correction lacks a controlled expansion or the required inclusion is absent, narrow the question rather than claiming a violation from an unproved entropy prescription. The deliverable remains useful even when it identifies a missing hypothesis instead of a new monotonicity theorem.

What is established for semiclassical horizons

Wall proves a rapidly changing-field GSL for arbitrary slices of a causal horizon by relating generalized entropy differences to monotonicity of relative entropy Wall 2012, arXiv v6, Theorem 1 and §§2–4. The proof avoids a quasistatic assumption, but it does not avoid its algebraic and semiclassical hypotheses. The exterior algebras must nest correctly, the reference state must have the required horizon properties, and the geometric constraint must convert modular-energy change into area change.

Faulkner and Speranza give an independent crossed-product formulation for arbitrary cuts of Killing horizons. For their semiclassical states, the algebra entropy agrees with generalized entropy and increases toward the future Faulkner and Speranza 2024, §3.3, Eqs. (3.57)–(3.58). Their construction also treats an algebra at infinity for interacting theories and discusses rotating/Unruh-state qualifications Faulkner and Speranza 2024, §§4.1–4.3.

Kirklin extends the framework perturbatively beyond the ordinary semiclassical limit using relational “dynamical cuts.” The modified inequality includes the free energy of the degrees of freedom that locate the later cut and reduces to the standard GSL only after additional semiclassical and energy assumptions Kirklin 2025, §§2–4. This is substantial progress, but “all orders in perturbative gravity” is not “an exact nonperturbative theorem.”

Results and open questions at 2026-08-20

Evidence	What it supports	What it does not establish
Wall’s relative-entropy theorem	Monotonic $S_{\mathrm{gen}}$ for the stated semiclassical causal-horizon setup, including rapid changes and arbitrary cuts	Trapping-horizon entropy, arbitrary higher-curvature/nonlocal gravity, or a nonperturbative Hilbert space
Crossed-product horizon algebras	A UV-finite algebraic entropy interpretation and a leading semiclassical GSL for Killing horizons	A sign for every subleading entropy correction
Relational dynamical cuts	A modified perturbative inequality beyond the usual fixed-cut semiclassical treatment	A unique physical cut, realistic clock in every setting, or exact completion
Classical area theorem	The $S_{\mathrm{out}}\rightarrow0$ limit under classical energy and predictability assumptions	Quantum fields, negative energy, or renormalized exterior entropy
Explicit subleading calculation	The first crossed-product correction can be written and bounded	Monotonicity of correction differences between finite cuts

The theorem rows outweigh analogies or examples because they identify every sign-bearing step. The last row is adverse rather than merely incomplete: the same paper that derives the correction says its contribution to the second-law difference has no proved sign Faulkner and Speranza 2024, §§5–7.

Strongest competing interpretations

A theorem-domain interpretation calls the GSL established wherever relative-entropy monotonicity, horizon nesting, the state, and the semiclassical constraint are controlled. This is the working assessment. A quasi-local interpretation expects a suitable entropy law for apparent or trapping horizons because those surfaces better track local dynamics. That programme may be physically valuable, but it changes the surface and usually the monotonicity mechanism; it is not covered by substituting one horizon word for another.

A fundamental-algebra interpretation treats generalized entropy as the von Neumann entropy of an exact gravitational algebra and expects a deeper data-processing theorem. Crossed products and relational cuts support this direction, while the unproved sign of subleading differences shows that the exact inference has not yet been earned. These positions disagree mainly about extrapolation, not about Wall’s theorem.

Counterexamples, null results, and failed routes

Replacing a causal horizon by a locally defined surface is a failed route to a free generalization: the exterior regions need not form the inclusion used by data processing, and area evolution can depend on the slicing. This does not refute the causal-horizon GSL; it removes a commonly claimed extension.

Likewise, bounding each subleading entropy correction does not bound the difference of corrections at two cuts with the needed sign. Faulkner and Speranza explicitly find the monotonicity question inconclusive. No accepted counterexample inside Wall’s hypotheses was found at the cutoff. The negative result is instead that the available algebraic calculation does not close the beyond-leading proof.

What would change the assessment

A proof that an exact or systematically corrected gravitational entropy obeys data processing for nested horizon algebras would broaden “established” beyond the leading semiclassical domain. A controlled counterexample satisfying Wall’s hypotheses would overturn the positive core. For quasi-local horizons, a theorem must state the surface prescription, state/energy conditions, renormalized entropy, and foliation dependence; examples of increasing entropy in selected simulations are not enough.

Research checklist

• Name the horizon and two cuts; do not use “horizon” as an interchangeable label.
• State the field algebra, reference state, asymptotics, and renormalization scheme.
• Identify the exact inclusion or quantum channel that supplies monotonicity.
• Derive the area/modular-energy relation at the same perturbative order as the entropy.
• Track edge, charge, and infinity algebras rather than dropping them from the subsystem.
• Test a nonstationary state and the classical limit, then isolate the first term whose sign is not controlled.
• If proposing a local horizon law, vary the foliation and show which conclusion is invariant.

No independent reproduction of the full proofs was completed for this assessment. The reproducibility prescription is analytic: trace every hypothesis of the named theorem to the proposed state and horizon, verify the nesting direction, and recompute the sign of each correction in one explicit nonstationary example.

Source and data map

Wall 2012 supplies the primary semiclassical theorem. Faulkner–Speranza 2024 supplies the independent crossed-product construction and its strongest internal limitation. Kirklin 2025 supplies the current perturbative extension with dynamical cuts. Wald 2001, §§3–4 fixes the distinction between classical area and generalized entropy. There is no observational dataset or public numerical product central to this record; source identity, version, theorem locator, and assumptions are the reproducibility-critical data.

References

• Faulkner, Thomas, and Antony J. Speranza. “Gravitational Algebras and the Generalized Second Law.” arXiv:2405.00847v2 [hep-th] (2024). arXiv.
• Kirklin, Josh. “Generalised Second Law Beyond the Semiclassical Regime.” Journal of High Energy Physics 07 (2025): 192. DOI. Open text.
• Wald, Robert M. “The Thermodynamics of Black Holes.” Living Reviews in Relativity 4 (2001): 6. DOI.
• Wall, Aron C. “A Proof of the Generalized Second Law for Rapidly Changing Fields and Arbitrary Horizon Slices.” Physical Review D 85 (2012): 104049. DOI. Open text, arXiv:1105.3445v6.