The full desk, newest first.
Every story Pulse24 is tracking, in chronological order — newest first.
Fable 5 Leads Optimization Task
Fable 5's consistent performance on an NP-hard problem highlights core model intelligence for complex optimisation. Architects and procurement teams should prioritise empirical validation over feature claims, as native guidance modes can degrade mean performance.
AI Doubles QLED Efficiency
AI-driven process design for QLEDs, developed by Seoul National and Sungkyunkwan Universities, doubles device efficiency and extends operational lifetime 40-fold. This breakthrough significantly cuts R&D time for next-generation displays, offering a path to more durable and energy-efficient products.
OpenAI Uses Minor Permits for Power
OpenAI's Stargate project used minor air permits to build a 360MW gas plant in Texas, bypassing environmental review. This mechanism, enabling 15 data centre-tied gas plants, locks in fossil fuel infrastructure and creates regulatory uncertainty for procurement teams.
AI Threats Boost Executive Security
Threats against AI executives surged sevenfold, forcing tech companies to significantly increase security budgets. This impacts operational costs and risk profiles for founders and procurement teams, shifting executive protection to a strategic imperative amid escalating AI-related risks.
Anthropic Pushes Stricter AI Regulation
Anthropic advocates for stricter state AI regulations, including mandatory third-party audits for large models, increasing compliance burdens for emerging developers. This follows a US order to suspend foreign AI access, signaling greater government intervention and requiring enhanced due diligence from procurement and platform teams.
Moonshot Unveils 2.8T-Parameter Kimi K3
China's Moonshot AI launched Kimi K3, a 2.8 trillion-parameter open-weight model, challenging US frontier AI. Its performance, validated by third-party benchmarks, shifts access to advanced AI capabilities towards customisable open-weight systems, impacting procurement and security teams.
AI Models Adopt State Censorship
A Meta Oversight Board study reveals major AI models refuse politically critical content about restrictive governments more often than permissive ones. This risks spreading state censorship globally, requiring procurement teams to assess model biases.
Google AI Mode Links Apps
Google's AI Mode now links directly with apps like Instacart and Canva, enabling task completion beyond simple queries. This shifts AI from information retrieval to agentic execution, directly affecting how applications interact with AI agents.
Cadence Launches AuraStack AI Agent
Cadence Design Systems launched AuraStack, an AI agent for PCB and chip package design, claiming it can halve time to market and boost productivity 15-fold. This extends AI automation across hardware development, impacting design economics and requiring assessment of consumption-based pricing.
OpenAI AI Red Team Strengthens GPT-5.6
Automated AI red-teaming scales vulnerability detection, shifting security efforts for platform engineers and security architects. OpenAI's GPT-Red achieved 84% success in finding prompt injection flaws in GPT-5.6, compared to 13% for human teams, demonstrating a mechanism for more robust model hardening.
Technion AI Predicts Heart Failure
Early heart failure detection enables preventive interventions, reducing hospitalisations and mortality. Technion's DeepHHF AI model predicts heart failure risk up to five years early from routine ECGs, offering health insurers and providers a cost-effective mechanism to target high-risk patients.
Apple Secures China AI Approval
Apple secured Chinese government approval for Apple Intelligence, integrating Alibaba and Baidu models. This opens a critical market for Apple's AI features, forcing platform engineers to adapt to local model integration for market access and boosting Apple's competitive position.
Suno Hack Reveals YouTube Scraping
A supply chain attack on AI music generator Suno revealed alleged scraping of YouTube data for model training, alongside a customer data breach. This raises significant legal and ethical questions for CTOs and procurement teams regarding data sourcing and vendor transparency in the AI sector.
Microsoft Patches Record 570 Flaws, Citing AI
Microsoft's record 570 security patches, driven by AI vulnerability discovery, signal an accelerated pace of required updates. This increased volume, including actively exploited zero-days, demands security and platform teams prepare for continuous, rapid patch management and resource allocation.
CXMT Prepares Asia's Largest IPO
CXMT's IPO, Asia's largest this year, strengthens China's domestic memory chip production, shifting global supply chain dynamics. Procurement teams face increased geopolitical influence on critical component access.
ASML Raises Forecast, Expands Capacity
ASML raised its 2026 revenue forecast by 16% and plans to expand EUV/DUV tool capacity by 30% annually, driven by AI chip demand. This impacts procurement teams and foundry operators, intensifying competition for critical chip manufacturing equipment.
AI Voice Fraud Defeats Detection
AI voice cloning has defeated traditional fraud detection, leading to over $893 million in reported losses in 2025. This shifts the defence burden to systemic institutional controls, as voice-cloning systems require minimal audio to create indistinguishable synthetic voices, rendering human detection unreliable.
Anthropic, Blackstone Fund Ode
Anthropic and Blackstone launched Ode with Anthropic, a $1.5 billion AI implementation company, acquiring Fractional AI. This signals frontier model developers are directly entering the enterprise integration market, intensifying competition for AI engineering talent and shifting how companies access advanced AI deployment expertise.
Codex Encrypts MultiAgentV2 Audit Trails
OpenAI's Codex CLI now encrypts MultiAgentV2 messages, removing human-readable audit trails for subagent tasks. This hinders debuggability and auditability for developers and security teams, increasing operational overhead and risk in agentic workflows.
OpenAI Researcher Launches Drug Startup
Top AI talent from OpenAI is shifting to drug discovery, with Miles Wang launching a startup seeking $200 million at a $2 billion valuation. This move highlights continued investor confidence and the potential for AI models to accelerate drug development and repurposing, offering faster routes to market.