More than 100 U.S. companies and government agencies have received authorization to use Mythos 5, with access extending to their non-American staff members as well.

Liquid AI, created by former MIT computer scientists, has introduced its smallest AI language model yet, LFM2.5-230M. This 230-million-parameter model is engineered for on-device agentic workflows, enabling it to run efficiently on smartphones, laptops, and robotics. Despite its small size, it surpasses larger models—more than four times its size—in data extraction tasks, outperforming Alibaba's 800M parameter Qwen3.5-0.8B and Google's 1B parameter Gemma 3 1B.

LFM2.5-230M focuses on developers and engineers building lightweight data extraction pipelines and edge systems. It operates under a dual-use commercial license, free for entities earning below $10 million annually, while larger enterprises require a paid license. The model’s unique LFM2 architecture combines gated short-range convolutions with grouped-query attention, allowing high inference speed and a memory footprint below 400MB. It supports a large 32K context window for processing long documents or data streams.

Performance benchmarks show the model's efficiency, decoding 213 tokens per second on a Samsung Galaxy S25 Ultra and 42 tokens per second on a Raspberry Pi 5. It excels specifically in AI ETL tasks, offering a cost-effective alternative to expensive cloud-based models by running locally. LFM2.5-230M scores notably well on benchmarks like BFCLv3 and CaseReportBench, demonstrating superior tool-use and data extraction performance compared to larger models.

The model also serves advanced research needs, demonstrated by its deployment on a Unitree G1 humanoid robot for on-device command processing and multi-step planning. The base and post-trained models are available on Hugging Face, with broad support for popular inference platforms.

Liquid AI released LFM2.5-230M under a dual-use license that is free for smaller users but restricts commercial use by companies earning over $10 million annually, requiring paid agreements for larger enterprises. This approach balances open accessibility with commercial protection.

OpenAI has unveiled the GPT-5.6 series, featuring three specialized models—Sol, Terra, and Luna—each tailored to different enterprise needs. Sol is designed for complex coding and cybersecurity tasks, Terra targets high-volume business workflows, and Luna is optimized for everyday, cost-efficient operations like summarization and drafting. Initially, these models will be accessible to roughly 20 select organizations following coordination with the U.S. government, in line with a recent executive order establishing AI safety benchmarks. OpenAI emphasizes a phased release to ensure security and compliance, introducing innovative features like extended reasoning modes and subagent workflows. Pricing varies, with Sol priced at $35 per million tokens, Terra at $17.50, and Luna at $7.00. The models demonstrate significant performance improvements across reasoning, biology, and cybersecurity tasks, supported by advanced safety measures and a revamped caching system for cost predictability. This release reflects the increasing intersection of AI innovation with national security frameworks, highlighting the challenges enterprises face in accessing cutting-edge AI tools amid regulatory scrutiny.

In 2026, "taste" has emerged as a defining concept in marketing—an elusive mix of sensitivity, intuition, and learned judgment that shapes how people and brands uniquely combine ideas, styles, and culture. Beyond algorithms and AI-driven sameness, Taste with a capital T stands out as a human-driven differentiator that reveals personal narrative, cultural literacy, and thoughtful curation. It requires curiosity, evolves with time, and balances saying yes and no amid shifting cultural moods. Similarly, talent—whether influencers, creators, or celebrities—remains a powerful business asset, but brands often miss the mark by focusing on reach instead of cultural relevance and resonance. Successful brands view talent not merely as a media channel but as a transfer of meaning that connects deeply with their cultural positioning. As the landscape shifts, brands need agency partners who don't just deliver polished campaigns, but who act as cultural guides—helping to navigate who to collaborate with, which spaces to engage, and how to build taste and talent as delicate, long-term muscles rather than quick fixes.

AI was initially seen as a tool to scale businesses by replacing human involvement — automate, remove the middleman, and boost efficiency and margins like SaaS models. However, companies implementing AI in real-world operations find the truth is quite the opposite: the more power AI holds, the closer teams need to be to their customers continuously.

This paradox emerges because AI shifts the nature of software. Instead of just running fixed workflows, AI must now interpret complex signals, adapt real-time, and understand context — which only humans embedded deeply in the customer environment can provide. Without this human context, AI produces noise rather than meaningful insights.

Trust is central when deploying AI live. Leaders question AI’s reliability in their unique environments, and these questions require human expertise that no AI model alone can answer. High-stakes scenarios, like cybersecurity, exemplify this need: is a flagged login an attack or legitimate user activity? Such nuances demand ongoing human-AI collaboration.

Top AI-driven companies are investing heavily in embedding expert teams within customer operations. These teams quickly translate real-world conditions into system adjustments, refining AI continuously. Internally, tearing down silos to bring engineers and operators side-by-side accelerates rapid decision-making and problem-solving.

Successful companies rebuild workflows around AI's strengths, invest in deep contextual knowledge, and treat trust itself as a product—earned through transparency, collaboration, and human accountability.

Ultimately, AI doesn't scale by removing humans; it scales by bringing people and technology closer to deliver smarter, trusted solutions.

AI-powered recruitment tools excel at sifting through large volumes of applications, but their strict filtering mechanisms can unintentionally exclude unconventional yet highly valuable candidates—those who could drive your business forward. It's crucial to recognize that relying solely on rigid algorithms may mean missing out on diverse talent that doesn't fit typical patterns but offers unique skills and perspectives.

CEO Sam Altman is reportedly reluctant to proceed with OpenAI's IPO if the company's valuation falls below $1 trillion. This has led to discussions about potentially delaying the public offering to secure a more favorable valuation.

DeepSeek is preparing to significantly expand its workforce across all departments as part of its strategy to strengthen its position against competitors in the AI industry. This recruitment drive coincides with the company's plans to secure approximately $7.4 billion in funding to support its growth ambitions.

PwC’s Will O’Brien discusses with SiliconRepublic.com the steps Irish businesses should take to boost their cybersecurity as the EU Presidency approaches, highlighting the increasing cyber threat landscape during this period.

Volkswagen is reportedly planning to cut about 100,000 jobs, which represents nearly 15% of its workforce, along with closing several plants in Germany. This move marks the largest restructuring in the automaker's history. The decision has sparked strong opposition from labor unions determined to resist these changes. The reduction comes as Europe's automotive sector contracts, with Volkswagen leading significant job cuts.