Large language models (LLMs) have significantly transformed natural language understanding and generation, but they raise privacy concerns due to potential exposure of sensitive information. Studies have highlighted the risk of information leakage, where adversaries can extract sensitive information embedded in the prompts. In this work, we introduce a novel private prediction framework for generating high-quality synthetic text with strong privacy guarantees. Our approach leverages the Differential Privacy (DP) framework to ensure worst-case theoretical bounds on information leakage without requiring any fine-tuning of the underlying models. The proposed method performs inference on private records and aggregates the resulting per-token output distributions. This enables the generation of longer and coherent synthetic text while maintaining privacy guarantees. Additionally, we propose a simple blending operation that combines private and public inference to further enhance utility. Empirical evaluations demonstrate that our approach outperforms previous state-of-the-art methods on in-context-learning (ICL) tasks, making it a promising direction for privacy-preserving text generation while maintaining high utility.

The verification of differential privacy algorithms that employ Gaussian distributions is little understood. This paper tackles the challenge of verifying such programs by introducing a novel approach to approximating probability distributions of loop-free programs that sample from both discrete and continuous distributions with computable probability density functions, including Gaussian and Laplace. We establish that verifying $(ε,δ)$-differential privacy for these programs is \emph{almost decidable}, meaning the problem is decidable for all values of $δ$ except those in a finite set. Our verification algorithm is based on computing probabilities to any desired precision by combining integral approximations, and tail probability bounds. The proposed methods are implemented in the tool, DipApprox, using the FLINT library for high-precision integral computations, and incorporate optimizations to enhance scalability. We validate {\ourtool} on fundamental privacy-preserving algorithms, such as Gaussian variants of the Sparse Vector Technique and Noisy Max, demonstrating its effectiveness in both confirming privacy guarantees and detecting violations.

Many synthesis and verification problems can be reduced to determining the truth of formulas over the real numbers. These formulas often involve constraints with integrals in them. To this end, we extend the framework of δ -decision procedures with techniques for handling integrals of user-specified real functions. We implement this decision procedure in the tool ∫dReal, which is built on top of dReal. We evaluate ∫dReal on a suite of problems that include formulas verifying the fairness of algorithms and the privacy and the utility of privacy mechanisms and formulas that synthesize parameters for the desired utility of privacy mechanisms. The performance of the tool in these experiments demonstrates the effectiveness of ∫dReal.

Security properties of real-time systems often in-volve reasoning about hyper-properties, as opposed to properties of single executions or trees of executions. These hyper-properties need to additionally be expressive enough to reason about real-time constraints. Examples of such properties include information flow, side channel attacks and service-level agreements. In this paper we study computational problems related to a branching-time, hyper-property extension of metric temporal logic (MTL) that we call HCMTL*. We consider both the interval-based and point-based semantics of this logic. The verification problem that we consider is to determine if a given HCMTL* formula ℑ is true in a system represented by a timed automaton. We show that this problem is undecidable. We then show that the verification problem is decidable if we consider executions upto a fixed time horizon$T$. Our decidability result relies on reducing the verification problem to the truth of an MSO formula over reals with a bounded time interval.

We consider the problem of checking the differential privacy of online randomized algorithms that process a stream of inputs and produce outputs corresponding to each input. This paper generalizes an automaton model called DiP automata [10] to describe such algorithms by allowing multiple real-valued storage variables. A DiP automaton is a parametric automaton whose behavior depends on the privacy budget ∈. An automaton A will be said to be differentially private if, for some D, the automaton is D∈-differentially private for all values of ∈ > 0. We identify a precise characterization of the class of all differentially private DiP automata. We show that the problem of determining if a given DiP automaton belongs to this class is PSPACE-complete. Our PSPACE algorithm also computes a value for D when the given automaton is differentially private. The algorithm has been implemented, and experiments demonstrating its effectiveness are presented.

Abstract A hyperproperty relates executions of a program and is used to formalize security objectives such as confidentiality, non-interference, privacy, and anonymity. Formally, a hyperproperty is a collection of allowable sets of executions. A program violates a hyperproperty if the set of its executions is not in the collection specified by the hyperproperty. The logic HyperCTL* has been proposed in the literature to formally specify and verify hyperproperties. The problem of checking whether a finite-state program satisfies a HyperCTL* formula is known to be decidable. However, the problem turns out to be undecidable for procedural (recursive) programs. Surprisingly, we show that decidability can be restored if we consider restricted classes of hyperproperties, namely those that relate only those executions of a program which have the same call-stack access pattern. We call such hyperproperties, stack-aware hyperproperties. Our decision procedure can be used as a proof method for establishing security objectives such as noninference for recursive programs, and also for refuting security objectives such as observational determinism. Further, if the call stack size is observable to the attacker, the decision procedure provides exact verification.

We introduce an automata model for describing interesting classes of differential privacy mechanisms/algorithms that include known mechanisms from the literature. These automata can model algorithms whose inputs can be an unbounded sequence of real-valued query answers. We consider the problem of checking whether there exists a constant d such that the algorithm described by these automata are dϵ-differentially private for all positive values of the privacy budget parameter ϵ. We show that this problem can be decided in time linear in the automaton's size by identifying a necessary and sufficient condition on the underlying graph of the automaton. This paper's results are the first decidability results known for algorithms with an unbounded number of query answers taking values from the set of reals.

Differential privacy is a mathematical framework for developing statistical computations with provable guarantees of privacy and accuracy. In contrast to the privacy component of differential privacy, which has a clear mathematical and intuitive meaning, the accuracy component of differential privacy does not have a generally accepted definition; accuracy claims of differential privacy algorithms vary from algorithm to algorithm and are not instantiations of a general definition. We identify program discontinuity as a common theme in existing ad hoc definitions and introduce an alternative notion of accuracy parametrized by, what we call, — the of an input x w.r.t. a deterministic computation f and a distance d , is the minimal distance d ( x , y ) over all y such that f ( y )≠ f ( x ). We show that our notion of accuracy subsumes the definition used in theoretical computer science, and captures known accuracy claims for differential privacy algorithms. In fact, our general notion of accuracy helps us prove better claims in some cases. Next, we study the decidability of accuracy. We first show that accuracy is in general undecidable. Then, we define a non-trivial class of probabilistic computations for which accuracy is decidable (unconditionally, or assuming Schanuel’s conjecture). We implement our decision procedure and experimentally evaluate the effectiveness of our approach for generating proofs or counterexamples of accuracy for common algorithms from the literature.

Differential privacy is a de facto standard for statistical computations over databases that contain private data. Its main and rather surprising strength is to guarantee individual privacy and yet allow for accurate statistical results. Thanks to its mathematical definition, differential privacy is also a natural target for formal analysis. A broad line of work develops and uses logical methods for proving privacy. A more recent and complementary line of work uses statistical methods for finding privacy violations. Although both lines of work are practically successful, they elide the fundamental question of decidability.

In recent years, a new approach has been developed for verifying security protocols with the aim of combining the benefits of symbolic attackers and the benefits of unconditional soundness: the technique of the computationally complete symbolic attacker of Bana and Comon (BC) [8]. In this article, we argue that the real breakthrough of this technique is the recent introduction of its version for indistinguishability [9], because, with the extensions we introduce here, for the first time, there is a computationally sound symbolic technique that is syntactically strikingly simple, to which translating standard computational security notions is a straightforward matter, and that can be effectively used for verification of not only equivalence properties but trace properties of protocols as well. We first fully develop the core elements of this newer version by introducing several new axioms. We illustrate the power and the diverse use of the introduced axioms on simple examples first. We introduce an axiom expressing the Decisional Diffie-Hellman property. We analyze the Diffie-Hellman key exchange, both in its simplest form and an authenticated version as well. We provide computationally sound verification of real-or-random secrecy of the Diffie-Hellman key exchange protocol for multiple sessions, without any restrictions on the computational implementation other than the DDH assumption. We also show authentication for a simplified version of the station-to-station protocol using UF-CMA assumption for digital signatures. Finally, we axiomatize IND-CPA, IND-CCA1, and IND-CCA2 security properties and illustrate their usage. We have formalized the axiomatic system in an interactive theorem prover, Coq, and have machine-checked the proofs of various auxiliary theorems and security properties of Diffie-Hellman and station-to-station protocol.

The design of security protocols is extremely subtle and vulnerable to potentially devastating flaws. As a result, many tools and techniques for the automated verification of protocol designs have been developed. Unfortunately, these tools don’t have the ability to model and reason about protocols with randomization, which are becoming increasingly prevalent in systems providing privacy and anonymity guarantees. The security guarantees of these systems are often formulated by means of the indistinguishability of two protocols. In this paper, we give the first practical algorithms for model checking indistinguishability properties of randomized security protocols against the powerful threat model of a bounded Dolev-Yao adversary. Our techniques are implemented in the Stochastic Protocol ANalayzer ( Span ) and evaluated on several examples. As part of our evaluation, we conduct the first automated analysis of an electronic voting protocol based on the 3-ballot design.

A probabilistic finite automaton (PFA) A is said to be regular-approximable with respect to (x,y), if there is a regular language that contains all words accepted by A with probability at least x+y, but does not contain any word accepted with probability at most x. We show that the problem of determining if a PFA A is regular-approximable with respect to (x,y) is not recursively enumerable. We then show that many tractable sub-classes of PFAs identified in the literature - hierarchical PFAs, polynomially ambiguous PFAs, and eventually weakly ergodic PFAs - are regular-approximable with respect to all (x,y). Establishing the regular-approximability of a PFA has the nice consequence that its value can be effectively approximated, and the emptiness problem can be decided under the assumption of isolation.

Model checking of systems formalized using probabilistic models such as discrete time Markov chains (DTMCs) and Markov decision processes (MDPs) can be reduced to computing constrained reachability properties. Linear programming methods to compute reachability probabilities for DTMCs and MDPs do not scale to large models. Thus, model checking tools often employ iterative methods to approximate reachability probabilities. These approximations can be far from the actual probabilities, leading to inaccurate model checking results. In this article, we present a new algorithm and its implementation that improves approximate results obtained by scalable techniques like value iteration to compute exact reachability probabilities.

We consider the problem of verifying the security of finitely many sessions of a protocol that tosses coins in addition to standard cryptographic primitives against a Dolev-Yao adversary. Two properties are investigated here - secrecy, which asks if no adversary interacting with a protocol P can determine a secret sec with probability > 1 - p; and indistinguishability, which asks if the probability observing any sequence 0̅ in P1is the same as that of observing 0̅ in P2, under the same adversary. Both secrecy and indistinguishability are known to be coNP-complete for non-randomized protocols. In contrast, we show that, for randomized protocols, secrecy and indistinguishability are both decidable in coNEXPTIME. We also prove a matching lower bound for the secrecy problem by reducing the non-satisfiability problem of monadic first order logic without equality.

Indistinguishability properties are essential in formal verification of cryptographic protocols. They are needed to model anonymity properties, strong versions of confidentiality, and resistance against offline guessing attacks. Indistinguishability properties can be conveniently modeled as equivalence properties. We present a novel procedure to verify equivalence properties for a bounded number of sessions of cryptographic protocols. As in the applied pi calculus, our protocol specification language is parametrized by a first-order sorted term signature and an equational theory that allows formalization of algebraic properties of cryptographic primitives. Our procedure is able to verify trace equivalence for determinate cryptographic protocols. On determinate protocols, trace equivalence coincides with observational equivalence, which can therefore be automatically verified for such processes. When protocols are not determinate, our procedure can be used for both under- and over-approximations of trace equivalence, which proved successful on examples. The procedure can handle a large set of cryptographic primitives, namely those whose equational theory is generated by an optimally reducing convergent rewrite system. The procedure is based on a fully abstract modelling of the traces of a bounded number of sessions of the protocols into first-order Horn clauses on which a dedicated resolution procedure is used to decide equivalence properties. We have shown that our procedure terminates for the class of subterm convergent equational theories. Moreover, the procedure has been implemented in a prototype tool Active Knowledge in Security Protocols and has been effectively tested on examples. Some of the examples were outside the scope of existing tools, including checking anonymity of an electronic voting protocol due to Okamoto.

Several measures have been proposed in literature for quantifying the information leaked by the public outputs of a program with secret inputs. We consider the problem of computing information leaked by a deterministic or probabilistic program when the measure of information is based on (a) min-entropy and (b) Shannon entropy. The key challenge in computing these measures is that we need the total number of possible outputs and, for each possible output, the number of inputs that lead to it. A direct computation of these quantities is infeasible because of the state-explosion problem. We therefore propose symbolic algorithms based on binary decision diagrams (BDDs). The advantage of our approach is that these symbolic algorithms can be easily implemented in any BDD-based model-checking tool that checks for reachability in deterministic non-recursive programs by computing program summaries. We demonstrate the validity of our approach by implementing these algorithms in a tool Moped-QLeak, which is built upon Moped, a model checker for Boolean programs. Finally, we show how this symbolic approach extends to probabilistic programs.

Information-theoretic measures based upon mutual information can be employed to quantify the information that an execution of a program reveals about its secret inputs. The information leakage bounding problem asks whether the information leaked by a program does not exceed a given threshold. We consider this problem for two scenarios: a) the outputs of the program are revealed, and b)the timing (measured in the number of execution steps) of the program is revealed. For both scenarios, we establish complexity results in the context of deterministic boolean programs, both for programs with and without recursion. In particular, we prove that for recursive programs the information leakage bounding problem is no harder than checking reachability.

For concurrent probabilistic programs having process-level nondeterminism, it is often necessary to restrict the class of schedulers that resolve nondeterminism to obtain sound and precise model checking algorithms. In this paper, we introduce two classes of schedulers called view consistent and locally Markovian schedulers and consider the model checking problem of concurrent, probabilistic programs under these alternate semantics. Specifically, given a B\"{u}chi automaton $Spec$, a threshold $x$ in $[0,1]$, and a concurrent program $P$, the model checking problem asks if the measure of computations of $P$ that satisfy $Spec$ is at least $x$, under all view consistent (or locally Markovian) schedulers. We give precise complexity results for the model checking problem (for different classes of B\"{u}chi automata specifications) and contrast it with the complexity under the standard semantics that considers all schedulers.

The main challenge in using abstractions effectively is to construct a suitable abstraction for the system being verified. One approach that tries to address this problem is that of counterexample guided abstraction refinement (CEGAR) , wherein one starts with a coarse abstraction of the system, and progressively refines it, based on invalid counterexamples seen in prior model checking runs, until either an abstraction proves the correctness of the system or a valid counterexample is generated. While CEGAR has been successfully used in verifying nonprobabilistic systems automatically, CEGAR has only recently been investigated in the context of probabilistic systems. The main issues that need to be tackled in order to extend the approach to probabilistic systems is a suitable notion of “counterexample”, algorithms to generate counterexamples, check their validity, and then automatically refine an abstraction based on an invalid counterexample. In this article, we address these issues, and present a CEGAR framework for Markov decision processes.

In this article, we introduce the model of finite state probabilistic monitors (FPM), which are finite state automata on infinite strings that have probabilistic transitions and an absorbing reject state. FPMs are a natural automata model that can be seen as either randomized run-time monitoring algorithms or as models of open, probabilistic reactive systems that can fail. We give a number of results that characterize, topologically as well as with respect to their computational power, the sets of languages recognized by FPMs. We also study the emptiness and universality problems for such automata and give exact complexity bounds for these problems.

The continuous run-time monitoring of the behavior of a system is a technique that is used both as a complementary approach to formal verification and testing to ensure reliability, as well as a means to discover emergent properties in a distributed system, like intrusion and event correlation. The monitors in all these scenarios can be abstractly viewed as automata that process a (unbounded) stream of events to and from the component being observed, and raise an ``alarm'' when an error or intrusion is discovered. These monitors indicate the absence of error or intrusion in a behavior implicitly by the absence of an alarm.In this paper we study the power of randomization in run-time monitoring. Specifically, we examine \emph{finite memory} monitoring algorithms that toss coins to make decisions on the behavior they are observing. We give a number of results that characterize, topologically as well as with respect to their computational power, the sets of sequences the monitors permit. We also present results on the complexity of deciding non-emptiness of the set of sequences permitted by a monitor.

We study a hybrid intuitionistic modal logic suitable for reasoning about distribution of resources. The modalities of the logic allow validation of properties in a particular place, in some place and in all places. We provide a sound and complete Kripke semantics. We also define a sound and complete birelational semantics, and show that it enjoys the finite model property: if a judgement is not valid in the logic, then there is a finite birelational counter-model. Hence, we prove that the logic is decidable.

Garay, Jakobsson and MacKenzie introduced the notion of abuse-free distributed contract-signing: at any stage of the protocol, no participant Ahas the ability to prove to an outside party, that A has the power to choose between completing the contract and aborting it. We study a version of this property, which is naturally formulated in terms of game strategies, and which we formally state and prove for a two-party, optimistic contract-signing protocol. We extend to this setting the formal inductive proof methods previously used in the formal analysis of simpler, trace-based properties of authentication protocols.