Category: Top News

  • Salesforce to Acquire the AI Customer Service Platform Fin (Intercom) for $3.6 Billion

    Salesforce to Acquire the AI Customer Service Platform Fin (Intercom) for $3.6 Billion

    IBL News | New York

    Salesforce said it’s acquiring the AI customer service platform Fin, formerly known as Intercom, for $3.6 billion.

    The deal, expected to close in the fourth quarter of the company’s fiscal year 2027, will result in the integration of Fin, with its 30,000 business customers, into the Salesforce Agentforce portfolio.

    Powered by its proprietary AI model, Apex, Fin’s AI Agent autonomously manages customer support queries across chat, email, phone, SMS, WhatsApp, and Slack.

    Salesforce CEO Marc Benioff in a release.

    Like many software-as-a-service companies, Salesforce is grappling with concerns that new AI tools will render its business model obsolete. The company’s shares have shed more than a third of their value in 2026.

    At the same time, the rise of agentic AI is heightening competition and forcing companies to beef up and invest in more autonomous tech for customers.

    Fin CEO Eoghan McCabe, who has publicly endorsed President Donald Trump and posted photos, started Intercom in 2011 with Des Traynor, Ciaran Lee, and David Barrett.

    In its three-decade history, Salesforce has made several multibillion-dollar acquisitions. Its largest was in 2021, the purchase of Slack for $27 billion.

     

  • Palantir-Nvidia Sovereign Deal Seeks to Reshape the AI Infrastructure Race

    Palantir-Nvidia Sovereign Deal Seeks to Reshape the AI Infrastructure Race

    IBL News | New York

    Governments and enterprises are increasingly showing they want AI they control, not rent, especially when trade secrets and national security sit on the line.

    Tapping into this demand, Palantir Technologies announced a partnership with Nvidia this week. The deal intends to integrate Nvidia’s Nemotron open-weight models into Palantir’s Sovereign AI Operating System.

    With this move, customers in U.S. government agencies and critical infrastructure can deploy, customize, and post-train models on their own data — while retaining full ownership of the resulting model weights.

    Palantir supplies the ontology layer that structures messy data into usable intelligence, as well as deployment tools via AIP, Foundry, and Apollo. Nvidia brings the hardware acceleration and open models. Together, they create a secure, on-premises or air-gapped stack that closed labs like OpenAI and Anthropic struggle to match in terms of data sovereignty.

    Palantir CEO Alex Karp [in the picture above] highlighted this exact point yesterday on a CNBC appearance. Enterprises and agencies grow tired of unpredictable token costs and the risk of transferring their competitive advantage to third parties. Palantir’s approach lets them own the means of production.

    Alex Karp’s message was clear: “Stop renting AI from Big Tech. Palantir and Nvidia just teamed up to give governments and giants complete control.”

    Palantir pitches from a position of strength. The company reported $1.6 billion in Q1 revenue, up 85% year-over-year — its fastest growth since going public. U.S. revenue hit $1.28 billion, more than double the year-ago period, with U.S. commercial up 133% and government up 84%.

    Sovereign AI infrastructure can grow into a $177 billion market by 2035 at a 28% CAGR, according to Precedence Research. Palantir doesn’t need to win every dollar — it only needs to become the default operating layer for the most sensitive workloads.

    Investors say the Palantir-Nvidia deal quietly strengthens Palantir’s position as an infrastructure provider rather than just another AI tool provider. With 85% revenue growth in Q1, guidance raised 71% for the full year, and a platform built for control-hungry customers, the setup favors patient shareholders who focus on execution over daily volatility.

  • Anthropic Restores Access on Fable 5 and Mythos 5 After Trump Administration Lifted Restrictions

    Anthropic Restores Access on Fable 5 and Mythos 5 After Trump Administration Lifted Restrictions

    IBL News | New York

    Anthropic will begin restoring access today to Claude Fable 5 and Mythos 5 after the Department of Commerce lifted export controls on its most advanced AI models.

    The move canceled a June 12 White House directive that banned all foreign nationals from accessing Anthropic’s flagship Fable 5 and Mythos 5 systems.

    The White House previously said it disabled the cutting-edge software worldwide because the San Francisco-based startup could not reliably filter out foreign users in real time.

    “Anthropic has agreed to proactively detect and address security risks associated with the models; to work diligently with the US government on protocols and standards and releases for Mythos, Fable, and future models; and to inform the U.S. government of any malicious activity,” Commerce Secretary Howard Lutnick wrote in a letter obtained by The New York Post.

    “The controls are withdrawn. A license is no longer required for the export, reexport, or in-country transfer, including deemed export or deemed reexport, of the Mythos or Fable models,” he added.

  • Anant Agarwal Transitions from edX to CEO of Grady, an AI Start-Up Specializing in Assessment and Feedback

    Anant Agarwal Transitions from edX to CEO of Grady, an AI Start-Up Specializing in Assessment and Feedback

    IBL News | New York

    Anant Agarwal, the founder of the edX platform, professor at MIT, and, since 2021, Chief Academic Officer at 2U, was hired as CEO of Grady, a Millburn, New Jersey-based, faculty-founded AI start-up specializing in grading and feedback. Agarwal will remain Founder and Senior Advisor at edX/2U.

    Grady holds every student to the instructor’s standard, and it keeps the instructor in charge. It effectively supercharges faculty and teaching assistants,” said Agarwal. 

    Founded by two university professors, Periklis Papakonstantinou of Rutgers University and Anastasios Sidiropoulos of the University of Illinois at Chicago, it is currently backed by Neotribe Ventures.

    “Every student deserves precise, fair feedback that helps them grow, whether in a ten-person seminar or a four-hundred-person course, at an Ivy or a community college,” explained Papakonstantinou.

    The instructor sets the standard, and Grady applies it across the course, so the same work is judged the same way whether it is the first assessment graded or the 80th. Students get detailed feedback while the material is still fresh, and faculty get back hours otherwise spent grading, along with a clearer view of the concepts a class has not yet grasped.

    In conversation with IBL News this week, Anant Agarwal [in the picture] said,

    “What drew me to Grady is the problem it takes on. Feedback and grading are incredibly important parts of learning and among the hardest to do well in a constrained educational system. At edX, we tried to build AI grading as far back as 2012, and I know firsthand how difficult it is. When I saw what Periklis and Anastasios had built, I was initially skeptical, so I tried it on my own MIT assignments. I was blown away! It graded handwritten work, schematics, and sketches, providing detailed feedback and requiring no training.”

    “I have long believed that AI has the potential to make every teacher a super teacher. Grady is the clearest example of that I have seen. It keeps the instructor fully in control, steering, reviewing, and editing the feedback and grades before they reach students, and it gives faculty their time back for the human parts of teaching that matter most. Used this way, AI amplifies the teacher rather than replacing them.”

    “This is an inflection point for education, much like the arrival of online learning was 15 years ago. I am energized to be building again and to help shape how AI strengthens teaching and learning in the years ahead.”

  • Anthropic Restores the Mythos Model to US Critical Infrastructure Operators

    Anthropic Restores the Mythos Model to US Critical Infrastructure Operators

    IBL News | New York

    Anthropic’s Mythos 5 is partially back, as about 100 trusted US organizations, primarily cybersecurity companies and critical infrastructure providers, do have access. Project Glasswing members, including Apple, Google, Cisco, Nvidia, and Microsoft, are authorized.

    • API users and international customers are still limited to older Claude models.
    • Fable 5, the general-use version, remains offline for everyone else.

    The company’s two most powerful AI models got pulled offline two weeks ago under a US government directive. Currently, Anthropic is actively negotiating to bring Fable 5 back for general use.

    Claude Mythos 5 is Anthropic’s strongest cybersecurity model. It can autonomously find software vulnerabilities at a scale, such as thousands of high-severity flaws across every major OS and browser, including a 27-year-old OpenBSD bug and chained Linux kernel exploits.

    That power is why it got locked down. The concern is that bad actors could use it to attack systems just as fast as defenders use it to protect them.

  • OpenAI is Reportedly Delaying Its IPO Until 2027

    OpenAI is Reportedly Delaying Its IPO Until 2027

    IBL News | New York

    OpenAI may delay its planned 2026 IPO (initial public offering) until next year, according to the New York Times,

    The San Francisco-based AI company is weighing whether to go public this year or hold off until 2027, as it aims to reach a $1 trillion valuation, up from the company’s last private valuation of $730 billion.

    OpenAI confidentially filed its S-1 paperwork with the SEC on June 8.

    Traders on the prediction market platform Kalshi think an OpenAI initial public offering will be announced by March 1, 2027.

    Previously, OpenAI was widely expected to go public in 2026, the same year as Anthropic’s.

    Also, at the beginning of June, OpenAI’s chief rival Anthropic confidentially filed for an IPO. Traders on Kalshi think there’s a 70% chance that Anthropic will officially announce a public market debut by December.

    The New York Times said SpaceX’s public market debut — the first of what was expected to be several mega-cap IPOs this year — has made OpenAI’s advisors more cautious.

    OpenAI has worried that Elon Musk’s company’s initial rally and subsequent fall signal that retail investors may have less interest in buying.

    But a cascade of recent developments has caused OpenAI’s executives to shift away from their most aggressive aspirations. Top of mind is what has happened to Elon Musk’s SpaceX after its I.P.O. this month. It was the largest ever, raising more than $85 billion and reaching a valuation of

    At its debut, SpaceX reached a valuation of $1.77 trillion, but since then the stock has been on a downward slide, with shares slumping to $153 at the end of Thursday’s trading day after reaching a high of $202 last week.

    Moreover, investors question now whether AI companies will live up to their sky-high promises.

    OpenAI reported roughly $13 billion in revenue in 2025, one of the people said, a number the company hopes to triple this year. OpenAI said this year that it was generating $2 billion in monthly revenue and has more than 2 million business customers.

    OpenAI, run by Sam Altman [in the picture above], is currently spending heavily on marketing and recruiting high-profile engineering talent from companies like Meta and Google. It is exploring other revenue streams, including dabbling in placing ads within ChatGPT and striking e-commerce deals with companies like Shopify and Stripe that would allow users to buy things from online stores within ChatGPT.

  • Spotify Will Launch a User Tool for Generating Audiobooks

    Spotify Will Launch a User Tool for Generating Audiobooks

    IBL News | New York

    This June, Spotify will introduce an ElevenLabs-powered AI tool for self-publishing audiobooks on an invite-only basis, in beta, within the Spotify for Authors platform.

    Authors will be allowed to redistribute their AI-generated audiobooks anywhere.

    The AI-powered audiobook generation won’t bind authors to an exclusive contract, meaning they are free to publish their generated audiobooks anywhere.

    To date, authors have been allowed to submit digitally narrated audiobooks to ElevenLabs, Spotify, and Google Play Books.

    To date, Spotify has clocked over 1 million Audiobook+ subscriptions and a catalog of 700,000 titles.

    This year, it also started a program for authors to sell physical books in the U.S. and the U.K.

    Spotify is expected to generate $100 million in annualized recurring revenue.

    Spotify is also expanding its “Spotify for Authors” platform to support 10 more languages, including French, Canadian French, German, Dutch, Latin American Spanish, Swedish, Finnish, Icelandic, Danish, and Norwegian.

    ElevenLabs released its own self-publishing platform for authors in 2025.

  • OpenAI Issues ‘GPT-5.5-Cyber’, with Capabilities Similar to Mythos and Fable 5

    OpenAI Issues ‘GPT-5.5-Cyber’, with Capabilities Similar to Mythos and Fable 5

    IBL News | New York

    OpenAI launched the full version of GPT-5.5-Cyber this week, a model designed to hunt down and patch software vulnerabilities, with capabilities similar to those of Anthropic’s Mythos and Fable 5. However, OpenAI’s cyber model never drew a ban.

    Researchers say the same vulnerability-hunting runs in GPT-5.5, Claude Opus, and other models.

    At the time of release, Anthropic said its capabilities were too dangerous to ship. It also spent the month fighting the White House.

    Meanwhile, OpenAI, well-connected in Washington, D.C., has said nothing during the launch. Moreover, it hired a co-author of Trump’s AI plan last week.

    We’re expanding Daybreak to help democratize patching vulnerable software at machine speed.

    To discover and generate patches for critical vulnerabilities⁠ in major browsers, network infrastructure, and operating systems such as FreeBSD and the Linux kernel, OpenAI has released:

    • An update to the Codex Security plugin.
    • GPT‑5.5‑Cyber, a model that sets a performance on CyberGym of 85.6% compared with 81.8% for GPT‑5.5. [See graphic above]
    • Daybreak Cyber Partner Program.
    • Patch the Planet⁠: an initiative founded with Trail of Bits in collaboration with HackerOne, Calif researchers, and maintainers to help widely used open-source projects move from findings to fixes.

    Frontier AI models have been increasingly accelerating the discovery of vulnerabilities. Now, defenders are overwhelmed with the number of vulnerabilities found. Patching vulnerabilities is a bottleneck now.

    Models can navigate large codebases, reason through attack paths, validate hypotheses, and surface security issues that might otherwise stay hidden. Defenders absolutely need access to these capabilities and tools to fix what models can find now before attackers do.

    Beyond vulnerability reports, the key is to validate the issue, understand its impact, develop and test a patch, coordinate disclosure, and help teams deploy the fix.

    OpenAI revealed that since it launched the Codex Security cloud in research preview in March, it has scanned over 30 million commits across more than 30,000 codebases; human reviewers have manually marked more than 70,000 findings as fixed, and over 500,000 findings have been automatically determined to be fixed.

  • Google Released Its Latest Open Model, Gemma 4 12B, Multimodal and Working Locally

    Google Released Its Latest Open Model, Gemma 4 12B, Multimodal and Working Locally

    IBL News | New York

    Google’s DeepMind has released its latest open model, agentic, a multimodal Gemma 4 12B, for local use on a laptop. The model can generate rich visual insights and fully functional webpages.

    With advanced coding capabilities, it can execute scripts on the fly for tasks such as data analysis. It can rewrite text through voice commands.

    Google AI Edge Gallery, the company’s local AI showcase app, available on macOS, allows users to explore Gemma.

    The giant company said that “the new MacOS desktop version runs 100% on-device across the entire feature set, ensuring a powerful, fully offline experience.”

     

     

     

  • “Not Many Jobs Will Go Away, and AI’s Impact Will Be on Skills,” Says Expert on a WBG Event

    “Not Many Jobs Will Go Away, and AI’s Impact Will Be on Skills,” Says Expert on a WBG Event

    IBL News | Madrid, Spain

    “Not many jobs will go away; AI’s impact will be on skills, but not on jobs,” said Matt Sigelman, President of The Burning Glass Institute, when delivering his latest labor market data research on how AI is restructuring work. “The changes in the software industry illustrate what’s to come.”

    Sigelman was prominently featured in the run-up to the 10th World Bank Group Global Education Conference: “Building Skills for Tomorrow’s Jobs”, which took place in Spain, from June 17–19, 2026.

    Over 400 leaders from higher education, industry, technology, and public policy attended three days of panels, fireside chats, and a dedicated AI Day at this conference, organized by The World Bank Group (WBG) and hosted across IE University’s campuses in Madrid and Segovia.

    In a parallel event, lifelong educator Paul LeBlanc, former President of SNHU and currently a Visiting Scholar and Special Advisor at the Harvard University Graduate School of Education (HGSE), presented his upcoming book, “Reclaiming Purpose: The University in the AI World,” which will be launched on July 23.

    LeBlanc said that “now there is a unique opportunity to get back to the university’s original mission.” He explained, “We are going to have precision learning, a highly precise model against an all-fit model.”

    • “Universities are the answer to the world’s problems, as they have the mission to make the world better.”
    • “We will move from the knowledge economy to the care economy.”
    • “No one is built with the AI speed.”
    • “Universities need to structure their data for creating a holistic student profile in this AI age. Now data is seated in silos. First, we need data together.”

    The Building Skills for Tomorrow’s Jobs” conference landed on a forceful message: the challenge of the AI era is skills, not jobs. The data show little evidence of mass employment displacement, but dramatic changes in what happens inside jobs.

    Drawing on a dataset of 65 million US career histories and billions of job postings, the Burning Glass Institute found that approximately 30% of the skills required for an average job have changed over the past decade. Competencies that once remained relevant for a decade or more can now become obsolete within a few years of graduation.

    “Only about 13% of credentials lead to meaningful wage gain,” stated Matt Sigelman, President of The Burning Glass Institute. “We need a nutrition label on credentials, so employers know what skills a worker has and what qualifies them for a particular job.”

    Participants interviewed by IBL News agreed that AI handles entry-level tasks well, which creates a temptation for employers to eliminate junior positions. They shared Sigelman’s calling to governments to work with industry to define credential standards tied to real economic outcomes — not enrollment numbers.

    Beyond the job market, the conference program touched upon at least seven key themes, according to IBL News’ analysis:

    1. AI and Automation — How AI is redefining job roles, skill requirements, and entry points across industries, not just technology.

    2. Generative AI in Education — How LLMs and digital platforms are transforming teaching, assessment, and institutional operations.

    3. Future Skills — Consensus that creativity, collaboration, adaptability, and digital fluency are now baseline requirements. Technical knowledge alone has a shorter shelf life than ever.

    4. Lifelong Learning — The degree-then-career model is dead. Continuous upskilling and reskilling are a necessity, especially for informal workers, care workers, and employees in shrinking sectors.

    5. Work-Integrated Education — Real-world job exposure, co-curricular activities, and experiential learning embedded in degree programs, not as optional extras.

    6. University–Industry Partnerships — The biggest actionable theme. Multiple panels argued that curricula must be co-designed with employers and updated in near real time, rather than on 5-year review cycles.

    7. Job Clustering — Emerging career pathways based on evolving skill requirements, moving beyond traditional job titles to transferable skill profiles.

    According to The World Bank, remarkable examples during the conference were:

    Spain: Santander Open Academy now offers hundreds of free AI and cybersecurity courses, reaching millions of learners in a single year — faster than any traditional curriculum process.

    Argentina: World Bank Group is supporting a university–mining company partnership project to create 10,000+ direct jobs and 50,000 indirect jobs by 2033.

    Malaysia: IFC is investing in Asia Pacific University of Technology and Innovation (APU) to scale tech and innovation programs across Southeast Asia.

    Philippines: The government is incorporating Google Career Certificates and AWS certifications into the national qualifications framework, as the BPO industry pivots to cloud and AI.

    Industry certifications: AWS, Microsoft, Google, Cisco, and Salesforce credentials were highlighted as signals of real productivity, often more current than academic degrees.

    IBM: Cited by the Burning Glass Institute as a company that made a concerted effort to remove degree requirements — reinstating them in some cases, but measuring and analyzing where skills-based hiring was working and where it wasn’t.

    Among the panelists, leaders from IE University and Camilo Jose Cela University reflected on how economic, technological, and geopolitical forces are simultaneously reshaping education, employability, and entrepreneurship. They highlighted the profound implications of AI and digitalization, particularly at a time when 1.2 billion young people in developing countries are expected to enter the workforce in the next 10–15 years.

    Manuel Muñiz, Rector, IE University:
    “Periods of profound economic, technological, and geopolitical transformation place extraordinary demands on societies and institutions. The responsibility of universities is not simply to respond to change, but to help people understand it, navigate it, and ultimately shape it.”

    Juan José Güemes Barrios, VP of Strategy and President of the Center for Entrepreneurship & Innovation, IE University:
    “AI is not replacing human talent; it is redefining its value. As machines take over routine tasks, uniquely human capabilities — judgment, creativity, and an entrepreneurial mindset — become the real competitive advantage. Our responsibility, as educational institutions, is to prepare people not for the jobs that exist today, but for the opportunities we are not yet able to imagine.”

    Nieves Segovia, President, SEK Education Group:
    “Universities must evolve from teaching-centered institutions into connection nodes within broader ecosystems of learning, innovation, and social development — bridging the academic world, the business sector, social organizations, media, and public administrations.” “Beyond transmitting knowledge, higher education institutions are called to facilitate connections, build trust, and contribute to collective responses to the great challenges of our time.”

     

    No formal signed commitments emerged or were published, but participants implied consensus on the idea that
    universities are becoming platforms, not just campuses. The conversation pointed toward institutions acting as ecosystem orchestrators — connecting academic, corporate, governmental, and civil society actors.