 I'm Andrew Morgan and this is a short video abstract version of my Asia Crypt 2022 presentation on concurrently composable, non-interactive secure computation. Non-interactive secure computation, or NISC, refers to two-party secure computation with minimal round complexity, that is, two single message rounds where only one party the receiver receives the output from the protocol. We focus on fully malicious security, where intuitively an adversary with full control over one of the parties, even having the ability to deviate arbitrarily from the protocol, should learn nothing about the honest party's input besides what they could infer from the output of the computation. We formalize this as simulation-based security, where an adversary's view of the real execution of the protocol should be indistinguishable from their view of an idealized execution, where messages are simulated independently of inputs while a trusted third party computes the functionality. In fact, NISC with malicious security in the plain model is well-known from standard assumptions, though we note that a relaxation to super polynomial time simulation is required to circumvent a classical impossibility result and achieve security with fewer than four rounds. In our work, however, we investigate the achievability of stronger notions of security, specifically two intuitive desiderata that are not provided by classical definitions of simulation-based security are composability, or are guaranteed at security of a composition of protocols can be based on the security of its components, and concurrency, a guarantee that security of a protocol still holds when many instances can be run concurrently and an adversary might try to adaptively use the views of certain instances to break others. The first notion of security to consider and provide both guarantees is the universal composability or UC framework introduced by Canetti in 2001, which requires security against an environment able to invoke many instances of a protocol and control communications and interactions of multiple corrupted parties in those instances. Furthermore, there exists an analog of super polynomial time simulation for UC security known as ANGEL-based, or oracle-aided, UC security that preserves both the composability and concurrency guarantees. Specifically, it allows the simulator and environment both access to a super polynomial helper oracle, or ANGEL, but otherwise restricts them to polynomial time. Of course, since UC security is strictly stronger than standalone security, it has historically been far more difficult to prove the existence of UC secure computation protocols, and especially of round efficient UC secure protocols. While the theoretical minimum of two rounds was achieved for standalone security even prior to our results, the best known protocols for concurrently secure two-party computation, even table and composability, run in two simultaneous broadcast rounds and with thus required three single message rounds, and the best known UC secure protocol runs in a large, unspecified constant number of rounds. Thus the question has remained open, can we achieve concurrently composable, non-interactive or two-round secure computation with malicious SPS security in the plain model? Our work answers this question in the affirmative. We construct and prove the security of NISC with Oracle aided UC security based on two underlying assumptions. A standalone SPS secure NISC protocol, which we augment to achieve UC security by using a non-interactive tag-based commitment scheme which satisfies CCA security, a strengthened notion of hiding analogous to chosen ciphertext security for encryption. In my full talk, I'll focus in presenting our main theorem, our construction of a UC secure NISC protocol, and discussing at a high level how the subcomponents combine to achieve the strengthened notion of security. Additionally, I'll discuss the requirement of non-interactive CCA secure commitments, which are thus far known only based in relatively complex assumptions, and highlight our second key result, which demonstrates that in fact non-interactive CCA secure commitments can be constructed based on UC secure NISC, hence showing that these commitments are both sufficient and necessary for UC secure NISC, or in other words, the assumptions for a construction are essentially minimal. While details of the construction and proof of the second result are out of scope for my talk, I'll conclude by noting that they, along with the complete construction and proof of the NISC protocol, can be found in our paper.