U.S. Secretary of State Marco Rubio traveled to Paris to meet with G7 foreign ministers from Britain, Canada, France, Germany, Italy, and Japan. The discussions centered on the Russia-Ukraine war, the Iran conflict, energy policy, and artificial intelligence. President Donald Trump's envoys Steve Witkoff and Jared Kushner are also scheduled to meet Ukrainian and Russian officials in Abu Dhabi. Rubio emphasized that postwar security guarantees for Ukraine and Europe remain a top priority for allies.
Red Hat launched OpenShift 5, its first major version upgrade in seven years, at KubeCon + CloudNativeCon America 2026. The platform now includes agentic troubleshooting that automatically investigates cluster events and proposes fixes. A built-in Model Context Protocol server makes OpenShift accessible to any AI agent. Organizations including NASA Jet Propulsion Laboratory and Bharti Airtel have already adopted OpenShift Virtualization.
Eric Schmidt, the former Google CEO, spoke with Caltech president Thomas F. Rosenbaum about artificial intelligence. He said AI could help address major challenges in climate change and healthcare, while also warning about job displacement and the need for clearer decision-making processes. Schmidt praised Caltech as a leader in AI research and called for better public understanding of the technology.
Australia is experiencing a tenfold increase in data center investment over three years, driven by AI spending, according to Assistant Governor Sarah Hunter from the Reserve Bank of Australia. Much of the investment goes toward imported machinery, which limits the direct benefit to GDP. Still, the growth is creating jobs in construction, engineering, and power supply while increasing demand for electricity and critical minerals.
Instagram introduced an Effect Builder bot that lets users create custom filters using AI prompts through a web interface. Users can upload photos to preview effects and ask the AI to refine or change results. Once a filter passes a safety review, Instagram makes it available for use in Stories and links back to the creator's profile.
A global pledge to use AI for environmental good at the COP31 climate summit in Antalya, Turkey is drawing criticism for lacking strong commitments. The pledge includes measures like using clean energy for data centers and improving transparency, but the United States is not on board. A study published in Nature found that AI may increase carbon emissions by boosting fossil fuel use more than it helps renewable energy.
Researchers studying how large language models annotate texts found that models rarely follow reassigned definitions across political-text datasets. The counterfactual audit showed that accuracy improvements do not always reflect better understanding of rules. Researchers recommend checking definitions, labels, demonstrations, and training data together when evaluating model performance.
New paper trail tools designed to protect authors from AI accusations may not be fully effective. Researchers found that AI-generated content can be manipulated to appear more human-like. The publishing industry continues to struggle with how to handle these challenges as AI becomes more prevalent in writing.
Companies have huge untapped value in unstructured data like emails, contracts, and scanned files, with 80 to 90 percent of enterprise data falling into this category and mostly ignored. IDC projects the global datasphere will reach 400 zettabytes by 2028. Experts like Andrew Ng say businesses will need to spend hundreds of millions on data re-architecture to become AI ready.
Nations and large enterprises are building AI gigafactories, massive centers designed to train next-generation AI models using tens or hundreds of thousands of compute nodes. Key planning steps include securing power and land, using liquid cooling, overbuilding networking, and deploying software orchestration before installing hardware. The European Commission is working to establish multiple such facilities, with HPE and AMD outlining these considerations in a sponsored architectural white paper.
Key Takeaways
- Marco Rubio met G7 foreign ministers in Paris to discuss the Russia-Ukraine war, Iran conflict, energy, and AI policy, with postwar security guarantees for Ukraine as a top priority.
- Red Hat launched OpenShift 5 with agentic troubleshooting, a Model Context Protocol server, and post-quantum cryptography; NASA JPL and Bharti Airtel are among adopters.
- Former Google CEO Eric Schmidt spoke at Caltech, praising AI's potential in climate and healthcare while warning about job loss and decision-making transparency risks.
- Australia's data center investment grew tenfold in three years due to AI spending, creating jobs but limiting GDP benefits from imported machinery.
- Instagram launched an Effect Builder bot allowing users to create custom AI-generated filters for Stories via a web interface.
- The COP31 climate pledge on AI in Antalya, Turkey was criticized as weak, with the U.S. not on board and a Nature study linking AI to increased carbon emissions.
- Research shows large language models rarely follow reassigned definitions, meaning accuracy gains do not necessarily reflect improved understanding of rules.
- Paper trail tools may not fully protect authors from false AI-writing accusations, as AI-generated content can be manipulated to appear more human-like.
- Between 80 and 90 percent of enterprise data is unstructured and mostly ignored; experts like Andrew Ng say businesses need hundreds of millions in data re-architecture to become AI ready.
- AI gigafactories require massive power, land, and cooling infrastructure, with HPE and AMD outlining key considerations and the European Commission planning multiple facilities.
Rubio to discuss Iran Ukraine AI with G7 allies in France
U.S. Secretary of State Marco Rubio will meet G7 foreign ministers in Paris on Monday. The Russia-Ukraine war, the Iran conflict, energy and artificial intelligence will be main topics. Rubio said postwar security guarantees for Ukraine and Europe remain a top ally concern. The trip is his first foreign venture since midterm elections. Allies will be eager to hear how President Donald Trump plans to proceed on each issue.
Rubio meets G7 allies in Paris amid global crises
U.S. Secretary of State Marco Rubio will meet counterparts from Britain, Canada, France, Germany, Italy and Japan in Paris. Two days of talks coincide with major protests across Europe. The Russia-Ukraine war, the Iran conflict, energy and artificial intelligence are among the main topics. President Donald Trump's first major personnel decision installed Stephen Miller as chief of staff. The G7 meeting is not expected to produce a joint communique.
Rubio discusses Iran Ukraine AI at G7 meeting in France
U.S. Secretary of State Marco Rubio will meet G7 counterparts in Paris with unresolved world crises on the agenda. The Russia-Ukraine war, the Iran conflict, energy and artificial intelligence are main topics. President Donald Trump's envoys Steve Witkoff and Jared Kushner are set to meet Ukrainian and Russian officials in Abu Dhabi this week. Rubio said postwar security guarantees for Ukraine and Europe have been a top concern. Qatari and other mediators are also set to meet with Rubio in Paris.
Red Hat adds agentic AI to OpenShift 5 and RHEL updates
Red Hat announced OpenShift 5, the first major version upgrade in seven years, with AI-powered troubleshooting and stronger security. The updated Lightspeed interface lets administrators use natural language to generate system performance visualizations. OpenShift 5 introduces agentic troubleshooting that automatically investigates cluster events and proposes remediation steps. The platform also includes an AI-driven upgrade planner. Red Hat Enterprise Linux updates include enhanced protection against software tampering and post-quantum cryptography.
Red Hat launches OpenShift 5 for agentic AI era
Red Hat announced OpenShift version 5.0 at KubeCon + CloudNativeCon America 2026 with major AI upgrades. The new version features a built-in Model Context Protocol server and enhanced LightSpeed AI agents. OpenShift 5 adds hardened multi-tenant isolation, automated certificate lifecycle management and post-quantum cryptography. The MCP server makes OpenShift available to any AI agent via a simple update. Organizations including NASA Jet Propulsion Laboratory and Bharti Airtel have adopted OpenShift Virtualization.
Eric Schmidt discusses AI promise and public concerns at Caltech
Eric Schmidt, the former Google CEO, spoke with Caltech president Thomas F. Rosenbaum about artificial intelligence. He said AI could help solve major problems like climate change and healthcare. He also warned about risks such as job loss and the need for clearer AI decision-making. Schmidt praised Caltech as a leader in AI research and called for better public understanding of the technology.
Australia sees big data center growth from AI investment
Assistant Governor Sarah Hunter from the Reserve Bank of Australia said AI investment is driving a data center boom in Australia. Data center investment grew tenfold in three years from a low starting point. Much of the spending goes toward imported machinery, which limits the benefit to Australia's GDP. Still, the growth creates jobs in construction, engineering, and power supply. The data center expansion also increases demand for electricity and critical minerals.
Instagram launches AI tool for creating custom filters
Instagram's Effect Builder bot lets users create new filters using AI prompts on the web. Users can upload photos to preview effects and ask the AI to change or improve the results. The tool can rewrite a simple prompt into detailed style instructions. Once a filter passes a safety review, Instagram makes it available for anyone to use in Stories and links back to the creator's profile.
UN climate pledge on AI falls short, critics say
A global pledge to use AI for environmental good at the COP31 climate summit in Antalya, Turkey is being called weak. The pledge includes measures like using clean energy for data centers and improving transparency. However, critics say it lacks strong commitments and the United States is not on board. A study in Nature found that AI may increase carbon emissions by boosting fossil fuel use more than it helps renewable energy.
Study examines how language models learn to annotate texts
Researchers studied how large language models annotate social-science texts and whether they truly follow definitions. The team created a counterfactual audit that swaps definitions between labels to test accuracy. Across three political-text datasets, the models rarely followed reassigned definitions. The study found that accuracy improvements do not always mean the models understand the rules better. Researchers say definitions, labels, demonstrations, and training data should all be checked together.
Paper Trail Tools May Not Fully Protect Authors From AI Accusations
Authors face growing concerns about being accused of using AI to write, even though they may not have. New paper trail tools can record an author's work process to prove originality. However, researchers found that AI-generated content can be manipulated to look more human-like. The publishing industry still struggles with how to handle these challenges. The future of writing remains uncertain as AI becomes more common.
Enterprises Have Huge Untapped Value Hidden in Unstructured Data
Companies have spent years organizing structured data like spreadsheets and records. Now AI makes it possible to unlock the value in unstructured data like emails, contracts, and scanned files. Between 80 and 90 percent of enterprise data is unstructured and mostly ignored. IDC projects the global datasphere will reach 400 zettabytes by 2028. Experts like Andrew Ng say businesses will need to spend hundreds of millions on data re-architecture to become AI ready.
8 Key Considerations for Building Massive AI Gigafactories
Nations and large enterprises are building AI gigafactories, massive centers designed to train next-generation AI models. These facilities use tens or hundreds of thousands of compute nodes and require power measured in megawatts or gigawatts. Key planning steps include securing power and land, using liquid cooling, overbuilding networking, designing for hardware failure, and deploying software orchestration before installing hardware. The European Commission is working to establish multiple such facilities. HPE and AMD outlined these considerations in a sponsored architectural white paper.
Sources
- Rubio will discuss Iran, Ukraine, AI and more with G7 allies in France
- Rubio will discuss Iran, Ukraine, AI and more with G7 allies in France
- Rubio will discuss Iran, Ukraine, AI and more with G7 allies in France
- Red Hat adds agentic AI operations, stronger security in OpenShift 5 and RHEL updates
- Red Hat Unfurls OpenShift Platform for the Agentic AI Era
- Eric Schmidt on AI's promise and public distrust
- Australia’s Data Centre Boom: How AI Investment Is Reshaping Economic Demand
- Instagram’s Effect Builder makes new filters from AI prompts
- The global pledge to use AI for climate good is a cruel joke
- Definitions, Label Names, or Demonstrations? Auditing What Governs LLM Annotation
- Authors Can Defend Against AI Allegations With ‘Paper Trail’ Tools. Is It Enough?
- The Ninety Percent Problem: Unlocking Value in Unstructured Data for Enterprises
- The industrialization of AI: 8 critical considerations for building an AI gigafactory
Comments
Please log in to post a comment.