AI News Archive: May 18, 2026 — Part 12
Sourced from 500+ daily AI sources, scored by relevance.
- Anthropic to brief global financial watchdog on cyber flaws exposed by Mythos
US tech group will discuss capabilities of its new AI model with members of the Financial Stability Board
- Anthropic to brief global financial regulators on cyber flaws found by Claude Mythos
Anthropic will brief leading finance ministries and central banks on vulnerabilities in the global financial system's cyber defenses that its new AI model Claude Mythos Preview has uncovered. The article Anthropic to brief global financial regulators on cyber flaws found by Claude Mythos appeared first on The Decoder .
- Anthropic to brief global financial watchdog on cyber flaws exposed by Mythos
US tech group will discuss capabilities of its new AI model with members of the Financial Stability Board
- Anthropic to share Mythos cyber flaw findings with global finance watchdog
Startup has declined to release Claude Mythos AI model publicly amid fears it could be used by hackers Business live – latest updates Anthropic is to brief the global finance watchdog on the implications of its Claude Mythos AI model, whose potential threat to cyber defences has alarmed experts. The US startup will discuss Mythos with the Financial Stability Board (FSB), which is chaired by the governor of the Bank of England, Andrew Bailey. Continue reading...
- Anthropic Lets Mythos Users Share Cyber Threats With Others
How to restrict access while still allowing users to share threat information is a major challenge facing AI companies.
- Anthropic to let partners share Mythos cybersecurity findings with others
Anthropic to let partners share Mythos cybersecurity findings with others Reuters
- Anthropic to brief global financial watchdog on cyber flaws exposed by Mythos, FT reports
Anthropic to brief global financial watchdog on cyber flaws exposed by Mythos, FT reports Reuters
- Global centres in India slow hiring as AI reshapes work, ANSR CEO says
Global capability centres in India are slowing hiring. Companies are cautious about global uncertainties and the rise of artificial intelligence. This cautiousness leads to reduced hiring plans. New entrants will drive growth with flexible workforces. Companies are experimenting with smaller teams. This approach allows for scaling up as needed.
- NextEra, Dominion want to create a massive power company as AI drives energy demand in the US
NextEra, Dominion want to create a massive power company as AI drives energy demand in the US Boston Herald
- NextEra, Dominion want to create a massive power company as AI drives energy demand in the US
NextEra, Dominion want to create a massive power company as AI drives energy demand in the US East Bay Times
- NextEra, Dominion want to create a massive power company as AI drives energy demand in the US
NextEra, Dominion want to create a massive power company as AI drives energy demand in the US The Mercury News
- NextEra to acquire Dominion Energy for $67 billion in deal driven by data center demand
The combined company will serve 10 million utility customers across Florida, Virginia and the Carolinas.
- Dominion Energy agrees to $67B sale to NextEra in deal driven by data center demand
The combined company will serve 10 million utility customers across Florida, Virginia and the Carolinas.
- NextEra to Buy Dominion Energy for $67 Billion. AI Data Centers Need a Lot of Electricity.
NextEra to Buy Dominion Energy for $67 Billion. AI Data Centers Need a Lot of Electricity. Barron's
- NextEra-Dominion Merger Creates a New Utility Giant for the AI Age
NextEra-Dominion Merger Creates a New Utility Giant for the AI Age Barron's
- NextEra bets $66.8B on AI power boom with Dominion Energy acquisition
NextEra Energy announced a $66.8 billion acquisition of Dominion Energy, expanding its reach into Northern Virginia’s booming AI data-center market.
- NextEra seeks Dominion to create massive power company as AI drives energy demand
NextEra Energy is seeking to acquire Dominion Energy in an all-stock deal valued at about $67 billion.
- NextEra, Dominion want to create a massive power company as AI drives energy demand in the US
NextEra, Dominion want to create a massive power company as AI drives energy demand in the US AP News
- NextEra, Dominion want to create a massive power company as AI drives energy demand in the US
NextEra Energy is seeking to acquire Dominion Energy in an all-stock deal valued at about $67 billion, creating a massive power company as the energy needs of artificial intelligence drive demand higher in the U.S. It is one of the biggest proposed mergers so far this year and would create the world’s biggest regulated electric utility business by market capitalization, the companies said on Monday.
- Pope and co-founder of Anthropic to launch pontiff’s AI encyclical on May 25
Pope and co-founder of Anthropic to launch pontiff’s AI encyclical on May 25 Boston Herald
- Pope and co-founder of Anthropic to launch pontiff’s AI encyclical
Pope and co-founder of Anthropic to launch pontiff’s AI encyclical The Mercury News
- Pope and co-founder of Anthropic to launch pontiff's AI encyclical on May 25
Pope Leo XIV and Anthropic's co-founder Christopher Olah are set to launch the pontiff’s first encyclical on May 25
- Pope Leo XIV to launch his first encylical, a document on artificial intelligence, with Anthropic's co-founder
The presence of Anthropic's Christopher Olah at the Vatican is significant, and suggests that the U.S. pope's position on AI will become a new flashpoint with the Trump administration.
- Pope and co-founder of Anthropic to launch pontiff’s AI encyclical on May 25
Pope and co-founder of Anthropic to launch pontiff’s AI encyclical on May 25 AP News
- Pope and co-founder of Anthropic to launch pontiff’s AI encyclical on May 25
Pope Leo XIV and Anthropic's co-founder Christopher Olah are set to launch the pontiff’s first encyclical on May 25.
- The Pope Is Hooking Up With a Co-Founder of Anthropic for Collab on AI
On God.
- Pope and co-founder of Anthropic to launch pontiff’s AI encyclical on May 25
Pope and co-founder of Anthropic to launch pontiff’s AI encyclical on May 25 Toronto Star
- Pope Leo to address rise of AI in first major text on May 25
Pope Leo to address rise of AI in first major text on May 25 The Straits Times
- Pope Leo to address rise of AI in first major text
Pope Leo to address rise of AI in first major text The Japan Times
- Pope Leo XIV presents first AI encyclical, Anthropic co-founder invited as guest speaker
Pope Leo XIV will present his first encyclical on artificial intelligence on May 25. Anthropic co-founder Christopher Olah has been invited as a guest speaker. The article Pope Leo XIV presents first AI encyclical, Anthropic co-founder invited as guest speaker appeared first on The Decoder .
- Pope Leo to address rise of AI in first major text on May 25
Pope Leo to address rise of AI in first major text on May 25 Reuters
- As chip industry chases AI, US national labs look to newcomers for supercomputers
America's nuclear weapons program faces a computing crunch. High-precision simulations for nuclear weapons are becoming difficult to power. This is because major chip makers are focusing on artificial intelligence. Sandia National Laboratories is now testing chips from new companies like NextSilicon. This ensures the nation's security mission can continue uninterrupted.
- Dell Adds 1,000 Clients for AI Gear, Targets Corporate Users
Dell Technologies Inc. said it added 1,000 customers for a key AI product line in the past quarter as the company tries to capture business from traditional large businesses taking on new artificial intelligence workloads.
- Dell Scales Up the AI Supply Chain to Meet Demand
Dell Technologies Inc. Chief Executive Officer Michael Dell and Nvidia CEO Jensen Huang talk about how they work together on Dell's AI Factory which uses servers with Nvidia chips, software and services. They speak to Bloomberg's Ed Ludlow at the Dell World conference in Las Vegas. (Source: Bloomberg)
- Nvidia's Huang Sees China Becoming a More Open Market
Nvidia CEO Jensen Huang talks about the trip he took to China with President Donald Trump and other American business leaders. Huang says he didn't talk to the Chinese directly about selling them H200 chips. Huang and Michael Dell speak to Bloomberg's Ed Ludlow at the Dell World conference in Las Vegas. (Source: Bloomberg)
- 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.
- 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...