Category: Top News

  • Anthropic Expects to Break Even In 2028 and OpenAI In 2030

    Anthropic Expects to Break Even In 2028 and OpenAI In 2030

    IBL News | New York

    Anthropic, maker of Claude, expects to break even in 2028, The Wall Street Journal reported yesterday.

    By contrast, OpenAI forecasts operating losses of $74 billion, or roughly three-fourths of revenue at that year, expecting to turn a profit in 2030.

    The maker of OpenAI expects to burn 14 times as much cash as Anthropic.

    With the goal of becoming a trillion-dollar company, OpenAI, which is always in constant fundraising mode, says it is investing significantly more in chips and data centers, and doling out more stock-based compensation to attract top researchers.

    “Demand for AI exceeds available compute supply today,” an OpenAI spokesman said.

    Recently, OpenAI, valued at $500 billion, signed a string of new computing deals with cloud and chip giants. Altman said on X that the deals put OpenAI on the hook for up to $1.4 trillion in commitments over the next eight years.

    Anthropic, valued at $183 billion, is focused on increasing sales among corporate customers, which account for about 80% of revenue. It avoids OpenAI’s costly forays into image and video generation, which require significantly more computing power. Anthropic’s AI models have also taken off among coders.

    Founded four years ago by Dario Amodei [in the picture, on the right side], a former Google researcher who left OpenAI, Anthropic is centered on selling its Claude chatbot to businesses.

    Microsoft is OpenAI’s largest cloud provider, while Amazon and Google are for Anthropic.

  • OpenAI Launches a New GPT‑5.1 Model, Improved on Intelligence and Communication Style

    OpenAI Launches a New GPT‑5.1 Model, Improved on Intelligence and Communication Style

    IBL News | New York

    OpenAI yesterday announced an upgrade to the GPT-5 series with the release of an enhanced GPT‑5.1 Instant, “warmer, more intelligent, and better at following instructions and conversations,” and a new GPT‑5.1 Thinking, “easier to understand and faster on simple tasks, more persistent on complex ones.”

    “GPT‑5.1 improves meaningfully on both intelligence and communication style,” said the company.

    “We’ve also improved instruction following, so the model more reliably answers the question you actually asked.”

    These personality settings will be applied across all models. The original Cynical (formerly Cynic) and Nerdy (formerly Nerd) options introduced earlier this year will remain available unchanged under the same dropdown in personalization settings.

    Updates made in personalization settings will take effect across all chats, including ongoing conversations.

  • Mayor-Elect Zohran Mamdani Faces the Challenge of NYC’s Public School System

    Mayor-Elect Zohran Mamdani Faces the Challenge of NYC’s Public School System

    IBL News | New York

    Fixing the New York City public school system will be an enormous challenge for the mayor-elect, socialist Democrat Zohran Mamdani, as he announced plans to recruit more teachers, expand support for homeless students, and pledge free childcare for all babies and toddlers, The New York Times reported.

    New York has the largest education system in the U.S., with 880,000 students enrolled, 1,500 schools, and an annual operating budget of $41 billion.

    Zohran Mamdani has been candid about education, acknowledging that he is still learning about this unwieldy system and will rely on guidance from advisers and the teachers’ union, which endorsed him.

    In this regard, given that Mamdani has to select a school’s chancellor by January, Michael Mulgrew, president of that union, the United Federation of Teachers, encouraged the new mayor to retain Melissa Aviles-Ramos, who has run the system since last fall. The union also has a strong relationship with Meisha Ross Porter, a former school chancellor who remains popular among education leaders.

    New York City was once home to more than 1.1 million public school students, but that number has fallen due to a decrease in birth rates and as families have moved to cities where it is far more affordable to raise children.

    Projections suggest that the system can lose tens of thousands of additional students in the coming years.

    Another controversy is that New York schools allow students to use bathrooms that align with gender ideology. Additionally, the system introduced a Black studies curriculum and social studies lessons focused on race, culture, and sexuality.

    In immigration, New York teachers have banded together to shield migrant children from ICE police.

    “This is exactly the type of school system that the Trump administration would target,” wrote The Times.

    In New York City, federal funding accounts for a relatively small portion — 6 percent — of the education budget. But even minor funding disruptions can create chaos for thousands of children and their families.

    Regarding performance, for decades, New York, like many cities, has consistently underperformed in literacy.

    More than half of Black and Latino children are not proficient on state tests. The percentage of students who struggle the most rose during the past 15 years.

  • OpenAI, Anthropic, and Microsoft Provide Millions to Train Teachers on AI

    OpenAI, Anthropic, and Microsoft Provide Millions to Train Teachers on AI

    IBL News | New York

    OpenAI, Anthropic, and Microsoft are providing millions of dollars for AI training to the American Federation of Teachers, the country’s second-largest teachers union.

    Their goal is to keep teachers relevant and help students use the technology wisely. The common goal is to train America’s future workforce.

    “We are preparing kids for the future. That is our primary job. And AI, like it or not, is part of our world,” say teachers.

    In exchange, the tech companies have an opportunity to make inroads into schools and win over students in the race for AI dominance.

    “There is no one else who is helping us with this. That’s why we felt we needed to work with the largest corporations in the world,” AFT President Randi Weingarten said. “We went to them — they didn’t come to us.”

    Under the arrangement announced in July, Microsoft is contributing $12.5 million to AFT over five years. OpenAI is providing $8 million in funding and $2 million in technical resources, and Anthropic has offered $500,000.

    With the funds, AFT plans to build an AI training hub in New York City that will offer virtual and in-person workshops for teachers. The goal is to open at least two more hubs and train 400,000 teachers over the next five years.

    Meanwhile, the National Education Association, the country’s largest teachers union, announced its own partnership with Microsoft last month. The company has provided a $325,000 grant to help the NEA develop AI trainings in the form of “microcredentials” — online trainings open to the union’s 3 million members, said to Reuters Daaiyah Bilal, NEA’s senior director of education policy. The goal is to train at least 10,000 members this school year.

    The unions own the intellectual property for the trainings, which cover safety and privacy concerns alongside AI skills.

    The Trump administration has encouraged private investment, recently creating an AI Education Task Force as part of an effort to achieve “global dominance in artificial intelligence.”

    The federal government urged tech companies and other organizations to foot the bill. So far, more than 100 companies have signed up.

    Beyond training teachers. Microsoft unveiled a $4 billion initiative to fund AI training and research, and to gift its AI tools to teachers and students. It includes the AFT grant and a program that will give all school districts and community colleges in Washington, Microsoft’s home state, free access to Microsoft CoPilot tools.

    Google says it will commit $1 billion to AI education and job-training programs, including free access to its Gemini for Education platform for U.S. high schools.

    In their sessions, teachers generate lesson plans using ChatGPT, Google’s Gemini, Microsoft CoPilot, and two AI tools designed for schools, Khanmigo and Colorín Colorado.

    They even find ways to engage bored learners with AI-graded classwork instantly, turning lesson plans into podcasts or online storybooks, or generating images and creating illustrated flashcards in English and Spanish to teach vocabulary.

  • Anthropic Will Roll Out Its Claude model to Cognizant’s 350,000 Employees

    Anthropic Will Roll Out Its Claude model to Cognizant’s 350,000 Employees

    IBL News | New York

    Anthropic announced it will roll out its flagship Claude model to Cognizant Technology Solutions’ 350,000 employees in one of its biggest enterprise deals yet.

    The deal comes at a time when Claude is ramping up efforts to sell to corporate clients, while its rival, OpenAI, is focusing on consumer-driven ChatGPT-related apps.

    San Francisco-based Anthropic stated that approximately 80% of its revenue is driven by corporate customers, and it has more than 300,000 business clients.

    Anthropic said it has a team of what it calls forward-deployed engineers—or staff embedded within enterprises to teach them how to use AI.

    Anthropic last month reached a deal with IBM and also announced that Deloitte’s over 470,000 employees will use its models.

    For Cognizant, a Teaneck, New Jersey–based professional services company, the deal with Anthropic is intended to enhance its software development capabilities, enabling the firm to transition from a system integrator to an AI builder and orchestrate multi-step workflows, with human oversight, across corporate functions, engineering, and delivery teams.

    Claude Code, along with the Model Context Protocol (MCP) and the Agent SDK, will be used to accelerate coding tasks, testing, documentation, and DevOps workflows.

    Cognizant said, “It will deploy Claude models, Claude Code, MCP, and Agent SDK with Cognizant’s software development and AI platforms to deliver:

    • Software engineering productivity: Deploying Claude and Claude Code with Cognizant Flowsource Platform (which connects tools and teams across the software development process) to accelerate coding tasks, testing, documentation, and DevOps workflows with MCP-based access to developer tools.
    • Legacy modernization: Combining Cognizant’s modernization frameworks with Anthropic’s code understanding and transformation capabilities to speed analysis and refactoring across large codebases.
    • Agentification: Using Cognizant Neuro AI Multi-Agent Orchestration (which builds and coordinates multiple AI agents working together) and Anthropic’s Agent SDK to design reusable, domain-specific agents and multi-agent systems that operate with explicit policies, approvals, and human-in-the-loop controls.
    • Industry solutions: Developing vertical solutions, beginning with Financial Services, leveraging Cognizant Agent Foundry (which helps enterprises build and deploy AI agents at scale) with Claude to embed agentic workflows into regulated, enterprise environments.
    • Responsible AI: Advancing practices for safe deployment, monitoring, and operations at scale, aligned to enterprise governance needs and open standards such as MCP.”

  • Coca-Cola Released An Improved AI-Generated Commercial for Christmas

    Coca-Cola Released An Improved AI-Generated Commercial for Christmas

    IBL News | New York

    Coca-Cola’s holiday ads have been enhanced with an upgraded dose of AI following last year’s debut of this technology, which drew criticism from creative professionals.

    San Francisco-based Silverside developed one of the two commercials for “Holidays Are Coming,” which will run throughout the season.

     

    The Wall Street Journal commented on the quality of the video ads:

    “The wheels of the red delivery trucks in Coke’s new commercials look as if they’re all turning, rather than gliding like some did last year. The shiny-faced, spaced-out humans of 2024 have ceded their place to an expanded host of critters, letting Coke dodge the “uncanny valley” where nearly real simulations of people wind up unsettling viewers.”

    Other advertisers have also utilized generative AI to achieve speed and cost efficiencies, despite some people’s distaste for the technology and its potential to render jobs in the creative industries.

    However, Generative AI ads require considerable work: a team of artists works frame by frame, often pixel by pixel, to refine and tweak images.

    According to the trade group Interactive Advertising Bureau, 30% of TV commercials, social videos, and online videos this year are being built or enhanced using generative AI tools, up from 22% in 2023.

    Last month, tech giant Google unveiled its first completely AI-generated ad spot.

     

  • OpenAI Announced 1M Business Customers and 800M Users Weekly

    OpenAI Announced 1M Business Customers and 800M Users Weekly

    IBL News | New York 

    OpenAI announced yesterday that it has 1 million business customers worldwide using ChatGPT for Work, either directly or through its developer platform.

    Organizations related to areas such as financial services, healthcare, and retail are among the most active. Consumer adoption is also at high rates, with 800 million users weekly.

    The San Francisco-based lab is set to generate $13 billion in revenue this year as it continues to expand sales.

    To support the enterprise acceleration, OpenAI launched a new wave of tools and integrations, such as:

    • Company knowledge consolidates all the context from connected apps (Slack, SharePoint, Google Drive, GitHub, Canva, Figma, Zillow, Spotify, among others) into ChatGPT.
    • Codex model for code generation, refactoring, and workflow automation.
    • AgentKit allows users to create and build enterprise agents.
    • Multimodal models, such as the Image Generation APISora 2, gpt-realtime, and Realtime API to build production voice agents.
    • Databricks has made OpenAI models available natively on its stack.
    • The Agentic Commerce Protocol (ACP) enables the creation of conversational commerce experiences in ChatGPT. Shopify, Etsy, Walmart, PayPal, and Salesforce are among the companies utilizing this protocol.

    OpenAI quoted a recent Wharton study to highlight that 75% of enterprises report a positive ROI, and fewer than 5% report a negative return. “When AI is deployed with the right use case and infrastructure, teams see real results,” said the firm.

    This week, too, OpenAI agreed to pay Amazon.com $38 billion for computing power in a multiyear deal. That marks the first partnership between the startup and the cloud company.

  • IBM Releases as Open Source Its ‘Granite 4.0 Nano’ AI Models that Can Run on the Browser

    IBM Releases as Open Source Its ‘Granite 4.0 Nano’ AI Models that Can Run on the Browser

    IBL News | New York

    IBM last week released, under the Apache 2.0 license, four new Granite 4.0 Nano models, designed to be highly accessible and well-suited for developers building applications on consumer hardware, without relying on cloud computing.

    With these models, IBM is entering a crowded and rapidly evolving market of small language models (SLMs), competing with offerings like Qwen3, Google’s Gemma, LiquidAI’s LFM2, and Mistral’s dense models in the sub-2B parameter space.

    With this release, IBM is positioning Granite as a platform for building the next generation of lightweight, trustworthy AI systems.

    The 350M variants can run comfortably on a modern laptop CPU with 8–16GB of RAM, while the 1.5B models typically require a GPU with at least 6–8GB of VRAM for smooth performance.

    This is a fraction of the size of their server-bound counterparts from companies like OpenAI, Anthropic, and Google.

    This Granite 4.0 Nano family includes four open-source models now available on Hugging Face:

    • Granite-4.0-H-1B (~1.5B parameters) – Hybrid-SSM architecture
    • Granite-4.0-H-350M (~350M parameters) – Hybrid-SSM architecture
    • Granite-4.0-1B – Transformer-based variant, parameter count closer to 2B
    • Granite-4.0-350M – Transformer-based variant

    Overall, the Granite-4.0-1B achieved a leading average benchmark score of 68.3% across general knowledge, math, code, and safety domains.

    For developers and researchers seeking performance without overhead, the Nano release means they don’t need 70 billion parameters to build something powerful.

  • Elon Musk’s xAI Launched Grokipedia, a New AI-Powered Encyclopedia

    Elon Musk’s xAI Launched Grokipedia, a New AI-Powered Encyclopedia

    IBL News | New York

    Elon Musk, last week, unveiled Grokipedia.com, his own version of Wikipedia, the crowdsourced online encyclopedia.

    The new project’s entries will be edited by xAI, Elon Musk’s artificial intelligence company.

    The website featured over 885,000 entries in just three days. This number took Wikipedia over half a decade to reach.

    Wikipedia, which debuted almost 25 years ago and now includes eight million human-written entries, has faced increasing criticism from conservatives in recent months.

    Elon Musk and his political allies have argued that the online encyclopedia is too “woke” and excludes conservative media outlets from its approved citations

    “Grokipedia will be a massive improvement over Wikipedia,” said Musk. “It will purge out the propaganda flooding Wikipedia.”

    “Wikipedia has achieved a dominant position. I hope Grokipedia challenges it and is able to fix that,” said David Sacks, the A.I. czar of the Trump administration and an investor in several of Musk’s companies, in an episode of his podcast this month. “But the easier path might just be for Wikipedia to stop blackballing and censoring conservative publications, rather than having to rebuild that whole thing from scratch.”

    Jimmy Wales, a co-founder of Wikipedia, said he is leading an internal working group focused on promoting neutral points of view and developing guidelines to encourage academic research.

    In addition, he said that he did not think AI could replace the site’s accuracy.

    Visits to its website have declined by 8% this year, while visits from automatic scrapers used by AI companies to harvest data have increased. AI-generated summaries by search engines and chatbots are also deterring users from visiting Wikipedia.

  • Educause Introduced the Winners of the 2025 Award Program

    Educause Introduced the Winners of the 2025 Award Program

    IBL News | Nashville, Tennessee 

    “In higher education, we face today skepticism and scrutiny, and current developments are testing us,” said Dr. John O’Brien, President and CEO at Educause, during the opening talk of the annual conference, which took place this week in Nashville, Tennessee. “The work we do, together, has never been more important.”

    John O’Brien also introduced the winners of the 2025 Award Program:

    Leadership Award:
    • Elias G. Eldayrie, Senior VP and CIO at University of Florida

    • Helen Norris, Former CIO and Vice President for Information Technology at Chapman University

    Organizational Culture Award:
    Liv Gjestvang, Vice President and CIO at Denison University

    Rising Star Award:
    Michael McGarry, Academic Technology Lead and LMS Administrator at California State University, Channel Islands

    Community Leadership Award:
    David Sherry, Former CIO at Princeton University

    The next day, the President and CEO at Educause, during the presentation of the 2026 Top 10 report, advocated for clarity, resilience, and connection among institutions as they navigate uncertainty.

    Also on Wednesday, in conversation with reporters (including IBL News), Dr. John O’Brien (pictured) said, “We are at an inflection point and educators want to come together.”