Large language models don't "learn"--they copy. And that could change everything for the tech industry. O n Tuesday, researchers at Stanford and Yale revealed something that AI companies would prefer to keep hidden. Four popular large language models--OpenAI's GPT, Anthropic's Claude, Google's Gemini, and xAI's Grok--have stored large portions of some of the books they've been trained on, and can reproduce long excerpts from those books. In fact, when prompted strategically by researchers, Claude delivered the near-complete text of,,, and, in addition to thousands of words from books including and .

Are there any conditions under which a generative model's outputs are guaranteed not to infringe the copyrights of its training data? This is the question of "provable copyright protection" first posed by Vyas, Kakade, and Barak (ICML 2023). They define near access-freeness (NAF) and propose it as sufficient for protection. This paper revisits the question and establishes new foundations for provable copyright protection -- foundations that are firmer both technically and legally. First, we show that NAF alone does not prevent infringement. In fact, NAF models can enable verbatim copying, a blatant failure of copy protection that we dub being tainted. Then, we introduce our blameless copy protection framework for defining meaningful guarantees, and instantiate it with clean-room copy protection. Clean-room copy protection allows a user to control their risk of copying by behaving in a way that is unlikely to copy in a counterfactual clean-room setting. Finally, we formalize a common intuition about differential privacy and copyright by proving that DP implies clean-room copy protection when the dataset is golden, a copyright deduplication requirement.

Understanding temporal relationships and accurately reconstructing the event timeline is important for case law analysis, compliance monitoring, and legal summarization. However, existing benchmarks lack specialized language evaluation, leaving a gap in understanding how LLMs handle event ordering in legal contexts. We introduce LexTime, a dataset designed to evaluate LLMs' event ordering capabilities in legal language, consisting of 512 instances from U.S. Federal Complaints with annotated event pairs and their temporal relations. Our findings show that (1) LLMs are more accurate on legal event ordering than on narrative texts (up to +10.5%); (2) longer input contexts and implicit events boost accuracy, reaching 80.8% for implicit-explicit event pairs; (3) legal linguistic complexities and nested clauses remain a challenge. While performance is promising, specific features of legal texts remain a bottleneck for legal temporal event reasoning, and we propose concrete modeling directions to better address them.

Rise of the'porno-trolls': how one porn platform made millions suing its viewers Instead, it was a subpoena. He had been sued in federal court for illegally downloading 80 movies. Some of the titles sounded cryptic - Do Not Worry, We Are Only Friends - or banal, like International Relations Part 2. Others were less subtle: He Loved My Big Ass, He Loved My Big Butt, and My Big Booty Loves Anal. Brown, who had spent decades investigating sex crimes, claimed he had never watched any of them. His years "dealing with pimping", he wrote in a court filing, left him "with no interest in pornography". He had been married for 40 years, he did not need to download Hot Wife, another title in the list.

Legal claims refer to the plaintiff's demands in a case and are essential to guiding judicial reasoning and case resolution. While many works have focused on improving the efficiency of legal professionals, the research on helping non-professionals (e.g., plaintiffs) remains unexplored. This paper explores the problem of legal claim generation based on the given case's facts. First, we construct ClaimGen-CN, the first dataset for Chinese legal claim generation task, from various real-world legal disputes. Additionally, we design an evaluation metric tailored for assessing the generated claims, which encompasses two essential dimensions: factuality and clarity. Building on this, we conduct a comprehensive zero-shot evaluation of state-of-the-art general and legal-domain large language models. Our findings highlight the limitations of the current models in factual precision and expressive clarity, pointing to the need for more targeted development in this domain. To encourage further exploration of this important task, we will make the dataset publicly available.

Large Language Models (LLMs) are increasingly explored for legal argument generation, yet they pose significant risks of manipulation through hallucination and ungrounded persuasion, and often fail to utilize provided factual bases effectively or abstain when arguments are untenable. This paper introduces a novel reflective multi-agent method designed to address these challenges in the context of legally compliant persuasion. Our approach employs specialized agents (factor analyst and argument polisher) in an iterative refinement process to generate 3-ply legal arguments (plaintiff, defendant, rebuttal). We evaluate reflective multi-agent against single-agent, enhanced-prompt single-agent, and non-reflective multi-agent baselines using four diverse LLMs (GPT-4o, GPT-4o-mini, Llama-4-Maverick-17b-128e, Llama-4-Scout-17b-16e) across three legal scenarios: "arguable", "mismatched", and "non-arguable". Results demonstrate that the reflective multi-agent approach excels at successful abstention by preventing generation when arguments cannot be grounded, improves hallucination accuracy by reducing fabricated and misattributed factors and enhances factor utilization recall by better using the provided case facts. These findings suggest that structured reflection within a multi-agent framework offers a robust method for fostering ethical persuasion and mitigating manipulation in LLM-based legal argumentation systems.

Melania Trump Used as'Window-Dressing' in Elaborate Memecoin Fraud, Legal Filing Claims The first lady of the United States became a pawn in an intricate memecoin scam that resulted in millions of dollars in losses, crypto investors have alleged. A cryptocurrency promoted in January by US first lady Melania Trump was part of a sophisticated fraud that "leveraged celebrity association and'borrowed fame' to sell legitimacy to unsuspecting investors," a new legal filing has alleged. In April, crypto investors brought a federal class action lawsuit against Benjamin Chow, cofounder of crypto exchange Meteora, and Hayden Davis, cofounder of crypto venture capital firm Kelsier Labs, among other defendants, accusing them of a multimillion-dollar fraud involving a single memecoin, $M3M3. Later, the plaintiffs filed an amended complaint, expanding the allegations to include racketeering activity. They claimed the pair had colluded to rig the market for $LIBRA, a coin promoted by Javier Milei, president of Argentina, which collapsed in value shortly after launch.

Large Language Models (LLMs) are increasingly used to generate user-tailored summaries, adapting outputs to specific stakeholders. In legal contexts, this raises important questions about motivated reasoning -- how models strategically frame information to align with a stakeholder's position within the legal system. Building on theories of legal realism and recent trends in legal practice, we investigate how LLMs respond to prompts conditioned on different legal roles (e.g., judges, prosecutors, attorneys) when summarizing judicial decisions. We introduce an evaluation framework grounded in legal fact and reasoning inclusion, also considering favorability towards stakeholders. Our results show that even when prompts include balancing instructions, models exhibit selective inclusion patterns that reflect role-consistent perspectives. These findings raise broader concerns about how similar alignment may emerge as LLMs begin to infer user roles from prior interactions or context, even without explicit role instructions. Our results underscore the need for role-aware evaluation of LLM summarization behavior in high-stakes legal settings.

The integration of artificial intelligence (AI) into the legal domain has revolutionized judicial processes, enabling tasks such as legal judgment prediction (LJP), case summarization, precedent retrieval, and automated legal research. Text extraction, the process of isolating seven semantically meaningful segments--referred to as blocks--from unstructured judicial decisions, is a cornerstone of these applications. These blocks include plaintiff demands, plaintiff arguments, defendant arguments, court evaluation of evidence, judicial reasoning steps, applicable legal norms, and court decision. Accurate extraction is critical, as errors can lead to misinterpretations of case facts, biased predictions, or inefficiencies in judicial workflows, potentially undermining justice delivery in high-stakes contexts. Evaluation metrics are essential for quantifying extraction quality, enabling iterative model improvements and ensuring reliability. Traditional metrics rely on annotated ground truth, which is resource-intensive to produce, particularly for legal texts characterized by verbose narratives, domain-specific terminology, and jurisdiction-specific nuances. The scarcity of annotated legal corpora has driven the development of unsupervised metrics that leverage intrinsic document properties, such as term frequencies, semantic coherence, and structural patterns. These metrics must capture surface-level accuracy, semantic fidelity, logical structure, and legal-specific elements like citations and temporal consistency, while addressing ethical concerns such as fairness and neutrality in AI-driven legal systems [1, 2].

AI audits play a critical role in AI accountability and safety. One branch of the law for which AI audits are particularly salient is anti-discrimination law. Several areas of anti-discrimination law implicate the "less discriminatory alternative" (LDA) requirement, in which a protocol (e.g., model) is defensible if no less discriminatory protocol that achieves comparable performance can be found with a reasonable amount of effort. Notably, the burden of proving an LDA exists typically falls on the claimant (the party alleging discrimination). This creates a significant hurdle in AI cases, as the claimant would seemingly need to train a less discriminatory yet high-performing model, a task requiring resources and expertise beyond most litigants. Moreover, developers often shield information about and access to their model and training data as trade secrets, making it difficult to reproduce a similar model from scratch. In this work, we present a procedure enabling claimants to determine if an LDA exists, even when they have limited compute, data, information, and model access. We focus on the setting in which fairness is given by demographic parity and performance by binary cross-entropy loss. As our main result, we provide a novel closed-form upper bound for the loss-fairness Pareto frontier (PF). We show how the claimant can use it to fit a PF in the "low-resource regime," then extrapolate the PF that applies to the (large) model being contested, all without training a single large model. The expression thus serves as a scaling law for loss-fairness PFs. To use this scaling law, the claimant would require a small subsample of the train/test data. Then, the claimant can fit the context-specific PF by training as few as 7 (small) models. We stress test our main result in simulations, finding that our scaling law holds even when the exact conditions of our theory do not.