AI News Archive: May 18, 2026 — Part 13
Sourced from 500+ daily AI sources, scored by relevance.
- TRACE: Trajectory Correction from Cross-layer Evidence for Hallucination Reduction
Hallucination correction is not a one-direction problem. We show that intermediate layers are neither uniformly more truthful than final layers nor uniformly less trustworthy. Yet hallucination reduction is usually instantiated through one fixed intervention form: contrast one layer against another,...
- FOL2NS: Generating Natural Sentences from First-Order Logic
Translating formal language into natural language is a foundational challenge in NLP, driving various downstream applications in semantic parsing, theorem validation, and question answering. In this study, we introduce First-Order Logic to Natural Sentence (FOL2NS), a neurosymbolic framework designe...
- iPOE: Interpretable Prompt Optimization via Explanations
Prompt optimization has often been framed as a discrete search problem to find high-performing and robust instructions for an LLM. However, the search result might not make it transparent why and where specific prompt changes lead to performance gains. This is in contrast to how humans are instructe...
- How Good LLMs Are at Answering Bangla Medical Visual Questions? Dataset and Benchmarking
Recent advancements in Large Language Models (LLMs) and Large Vision Language Models (LVLMs) have enabled general-purpose systems to demonstrate promising capabilities in complex reasoning tasks, including those in the medical domain. Medical Visual Question Answering (MedVQA) has particularly benef...
- A Data-Efficient Path to Multilingual LLMs: Language Expansion via Post-training PARAM$Δ$ Integration into Upcycled MoE
Expanding Large Language Models~(LLMs) to new languages is a costly endeavor, demanding extensive Continued Pre-Training~(CPT) and data-intensive alignment. While recent data-free merging techniques attempt to bypass alignment by fusing a multilingual CPT-enhanced model with its instruct counterpart...
- MA$^{2}$P: A Meta-Cognitive Autonomous Intelligent Agents Framework for Complex Persuasion
Persuasive dialogue generation plays a vital role in decision-making, negotiation, counseling, and behavior change, yet it remains a challenging problem. In complex persuasion where the persuadee's internal states are not expressed clearly, the persuader must interpret responses, infer the persuadee...
- Monitoring the Internal Monologue: Probe Trajectories Reveal Reasoning Dynamics
Large Reasoning Models (LRMs) introduce new opportunities for safety monitoring through their Chain of Thought (CoT) reasoning. However, CoT is not always faithful to the model's final output, undermining its reliability as a monitoring tool. To address this, we investigate the hidden representation...
- What is Holding Back Latent Visual Reasoning?
Humans can approach complex visual problems by mentally simulating intermediate visual steps, rather than reasoning through language alone. Inspired by this, several works on Vision-Language Models have recently explored chain-of-thought reasoning with continuous latent tokens as intermediate visual...
- Machine Unlearning for Masked Diffusion Language Models
Recent masked diffusion language models (MDLMs), such as LLaDA and Dream, have achieved performance comparable to autoregressive large language models. Unlike autoregressive models, which generate text sequentially, MDLMs generate text by iteratively denoising masked positions in parallel. During fi...
- SIREM: Speech-Informed MRI Reconstruction with Learned Sampling
Real-time magnetic resonance imaging (rtMRI) of speech production enables non-invasive visualization of dynamic vocal-tract motion and is valuable for speech science and clinical assessment. However, rtMRI is fundamentally constrained by trade-offs among spatial resolution, temporal resolution, and ...
- Leveraging Graph Structure in Seq2Seq Models for Knowledge Graph Link Prediction
We introduce Graph-Augmented Sequence-to-Sequence (GA-S2S), a novel framework that integrates a T5-small encoder-decoder with a Relational Graph Attention Network (RGAT) to improve link prediction in knowledge graphs. While existing Seq2Seq models rely solely on surface-level textual descriptions of...
- Scalable Environments Drive Generalizable Agents
Generalizable agents should adapt to diverse tasks and unseen environments beyond their training distribution. This position paper argues that such generalization requires environment scaling: expanding the distribution of executable rule-sets that agents interact with, rather than only increasing t...
- SafeDiffusion-R1: Online Reward Steering for Safe Diffusion Post-Training
Diffusion models have been widely studied for removing unsafe content learned during pre-training. Existing methods require expensive supervised data, either unsafe-text paired with safe-image groundtruth or negative/positive image pairs, making them impractical to scale. Furthermore, offline reinfo...
- MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agents
Recent GUI agents have made substantial progress in visual grounding and action prediction, yet they remain brittle in long-horizon tasks that require maintaining task state across many interface transitions. Existing agents typically rely on raw history replay or text-only memory, which either over...
- Meta Moves 7,000 Workers Into AI Roles Ahead of Job Cuts
Meta Platforms Inc. is reassigning 7,000 workers to new jobs related to artificial intelligence, according to an internal memo, part of a broad corporate restructuring that includes planned staff reductions later this week.
- Pope Leo and co-founder of Anthropic to launch pontiff’s AI encyclical on May 25
Pope Leo and co-founder of Anthropic to launch pontiff’s AI encyclical on May 25 Chicago Tribune
- Amazon Alexa Plus can now create AI-generated podcasts
Alexa Plus, Amazon's upgraded AI assistant, can now generate podcasts on "virtually any topic," according to an announcement on Monday. With the update, Amazon says you can give Alexa Plus a topic, and the AI assistant will offer an overview of what its AI hosts plan to talk about, allowing you to steer the conversation […]
- Alexa+ Can Now Turn Any Topic You Suggest Into an AI-Generated Podcast
Alexa+ Can Now Turn Any Topic You Suggest Into an AI-Generated Podcast PCMag Australia
- Amazon Partners With Hundreds of News Outlets to Create Personalized AI-Generated Podcasts
Any topic can have a podcast within minutes.
- Amazon’s new Alexa+ powered feature can generate podcast episodes
Amazon’s Alexa+ can now generate custom AI podcasts on demand, as the company expands its assistant into a personalized AI content platform.
- Musk loses blockbuster OpenAI suit as jury says too late
A federal jury ruled Monday that Elon Musk waited too long to sue OpenAI and its co-founders, delivering a decisive victory to the ChatGPT startup and ending one of Silicon Valley's most closely watched courtroom battles.
- Jury to decide fate of Musk's blockbuster suit against OpenAI
Deliberations begin Monday in the blockbuster trial pitting Elon Musk against AI giant OpenAI and its CEO Sam Altman, whom Musk accuses of abandoning the company's founding mission.
- Federal court rejects Elon Musk's claims against OpenAI, saying he filed his lawsuit too late
Federal court rejects Elon Musk's claims against OpenAI, saying he filed his lawsuit too late
- Federal court rejects Elon Musk’s claims against OpenAI, saying he filed his lawsuit too late
Federal court rejects Elon Musk’s claims against OpenAI, saying he filed his lawsuit too late Boston Herald
- Jury rules against Elon Musk in his feud with OpenAI
Jury rules against Elon Musk in his feud with OpenAI Austin American-Statesman
- Jury rules against Elon Musk in his feud with OpenAI, saying he filed his lawsuit too late
Jury rules against Elon Musk in his feud with OpenAI, saying he filed his lawsuit too late The Mercury News
- Federal court rejects Elon Musk’s claims against OpenAI, saying he filed his lawsuit too late
Federal court rejects Elon Musk’s claims against OpenAI, saying he filed his lawsuit too late The Mercury News
- Elon Musk loses landmark trial against OpenAI because he waited too long
A nine-person jury ruled against Elon Musk in a high-stakes trial against OpenAI.
- A jury took less than two hours to completely crush Elon Musk’s AI suit
A jury took less than two hours to completely crush Elon Musk’s AI suit Newsweek
- Elon Musk Waited Too Long to Sue OpenAI and Sam Altman, Jury Rules
The tech billionaire missed statute of limitations deadlines on claims that OpenAI — which Musk co-founded as a nonprofit — breached its charitable duties and unjustly enriched its founders
- Sam Altman May Win Against Elon Musk in Court, but Will His Public Image Survive the Trial?
Musk’s lawyers may not have presented a winning argument, Tom Dotan writes, but at the end of this trial, it seems clear that the OpenAI nonprofit is little more than the most well-funded fig leaf of all time. Will anyone care?
- Elon Musk loses his case against Sam Altman
After around two hours of deliberation, the jury has reached a unanimous verdict in Musk v. Altman, the tech trial of the year. The group found that two claims were barred by the statute of limitations, and a third failed thanks to the dismissal of one of these. The jury here is an advisory jury, […]
- Elon Musk just lost his OpenAI lawsuit — and the fallout could reshape the AI race
Elon Musk just lost his OpenAI lawsuit — and the fallout could reshape the AI race Tom's Guide
- The 5 most unhinged revelations from Elon Musk’s lawsuit against OpenAI
Friendship breakups are never easy, but few are as messy and expensive as the collapse of Elon Musk and Sam Altman’s once thriving tech bromance, which has — for now — reached a legal end. On Monday, a jury ruled against Musk in his lawsuit against OpenAI, which contended that Altman and other executives “stole […]
- Sam Altman and OpenAI Beat Elon Musk in Court, Paving the Way for a Potential IPO
Musk's suit alleged multiple defendants colluded to steal a charity, corrupting a technological nonprofit from within.
- OpenAI’s courtroom win over Elon Musk clears a major obstacle to an IPO
OpenAI has prevailed in its fight against Elon Musk. A jury on Monday found that Musk did not file his lawsuit against the AI giant within the statute of limitations. The judge quickly agreed with the jury, making the ruling final. The win for OpenAI came after less than two hours of jury deliberation. Within 20 minutes, the judge, who could have taken up to a month to issue a final ruling in the case, agreed with the advisory jury and issued the final say. Musk had alleged that OpenAI “stole a charity” when it converted into a for-profit company. With the case now behind it, a major obstacle in OpenAI’s path toward becoming a publicly traded company has been cleared. “We’re in and out in under a month, and now OpenAI has a road to IPO,” Alex Kantrowitz, host of the Big Technology podcast, told CNBC. OpenAI has been working behind the scenes for some time to plan for a public offering that would likely be among the largest in Wall Street history. The company is currently valued at $500
- Musk v. Altman: Federal jury sides with OpenAI in legal battle between the 2 tech billionaires
A federal jury has sided with OpenAI and its top executives in a feud with Elon Musk, who accused them of betraying a shared vision for it to guide artificial intelligence ’s development as a nonprofit dedicated to humanity’s benefit. The nine-person jury unanimously found that Musk waited too long to file his lawsuit (Musk v. Altman et al.) and missed the deadline for the statute of limitations. Musk, the world’s richest man, was a co-founder of OpenAI, the company that launched in 2015 and went on to create ChatGPT. After investing $38 million in its first years, Musk accused OpenAI CEO Sam Altman and his top deputy of shifting into a moneymaking mode behind his back. The jury served in an advisory role, but Judge Yvonne Gonzalez Rogers accepted the verdict Monday as the court’s own and dismissed Musk’s claims. The trial that began April 27 in Oakland, California, shed light on the bitter falling-out between the two Silicon Valley titans and the beginnings of OpenAI, now a company va
- Elon Musk Just Lost His $150 Billion Lawsuit Against Sam Altman and OpenAI — Here's What Happens Next
Elon Musk Just Lost His $150 Billion Lawsuit Against Sam Altman and OpenAI — Here's What Happens Next entrepreneur.com
- Elon Musk has lost his case against OpenAI and Sam Altman
Elon Musk has lost his case against OpenAI and Sam Altman
- Jury Unanimously Sides With OpenAI, Tossing Out Elon Musk's Claims
A jury unanimously tossed out Elon Musk's claims against OpenAI on Monday, saying the statute of limitations had expired. The post Jury Unanimously Sides With OpenAI, Tossing Out Elon Musk's Claims appeared first on Investor's Business Daily .
- Jury To Decide Fate Of Musk's Blockbuster Suit Against OpenAI
Jury To Decide Fate Of Musk's Blockbuster Suit Against OpenAI Barron's
- FlintLab Sirius Platform
AI-Native Unified Device Infra PaaS
- OpenAI Found Not Liable in Elon Musk Suit
OpenAI Found Not Liable in Elon Musk Suit Barron's
- Federal jury delivers verdict on Musk's lawsuit against OpenAI
A jury in federal court reached a verdict in Elon Musk's lawsuit against OpenAI after Musk accused the maker of ChatGPT of violating its founding mission as a nonprofit.
- Jury throws out Elon Musk’s lawsuit against OpenAI and Sam Altman in less than two hours
The jury ruled that Elon Musk had waited too long to sue OpenAI and its other co-founders.
- Jury unanimously dismisses Elon Musk's lawsuit against OpenAI
California jury dismissed all charges, finding that Musk missed the three-year statute of limitations to file suit.
- Federal court rejects Elon Musk's claims against OpenAI, saying he filed his lawsuit too late
The trial shed light on the bitter falling-out between the two Silicon Valley titans and the beginnings of OpenAI, now a company valued at $852 billion and moving toward potentially one of the largest initial public offerings in history.
- Jury throws out Musk's lawsuit against OpenAI and Sam Altman
A federal judge in California dismissed a lawsuit from Elon Musk against the maker of ChatGPT after a jury found that he exceeded the statute of limitations. Musk had claimed that OpenAI betrayed its original mission of developing artificial intelligence for the good of humanity when it added a for-profit arm. Geoff Bennett discussed more with Gerrit De Vynck of The Washington Post.
- Pope Leo to launch encyclical on AI with Anthropic co-founder
The presence of Anthropic at the Vatican suggests that the pope's position on AI will become a new flashpoint with the Trump administration
- Jury dismisses all claims in Elon Musk's lawsuit against OpenAI CEO Sam Altman
Musk had sought to oust Altman from his leadership position over claims that he and others breached their duty to OpenAI's original nonprofit mission and unjustly enriched themselves.