The open-source AI movement has never lacked for options. Mistral, Falcon, and a growing field of open-weight models have been available to developers for years. But when Meta threw its weight behind Llama, something shifted. A company with three billion users, vast compute resources, and the credibility of a tech giant was now building openly, and the developer community responded.

By early 2026, the Llama ecosystem had reached 1.2 billion downloads, averaging about 1 million per day. That is the context for what happened on April 8, 2026. Meta launched Muse Spark, its first major new Meta AI model in a year, and the first product from its newly formed Meta Superintelligence Labs.

It is capable in ways Llama 4 never was, benchmarks well against the current frontier, and is completely proprietary. No free download. No open weights. No building on it unless Meta decides you can.

The companyspentUS$14.3 billion, brought in Alexandr Wang from Scale AI to lead its AI rebuild, then spent nine months tearing down its entire AI stack and starting over. Muse Spark is what came out the other side. The developer community that made Llama what it was is now being asked to wait for a future open-source version that may or may not arrive on any predictable timeline.

What is Muse Spark?

Muse Spark is a natively multimodal reasoning model with tool-use, visual chain of thought, and multi-agent orchestration built in. It now powers Meta AI, which reaches over three billion users in Meta’s apps. Meta rebuilt its technology infrastructure from scratch, letting the company create a model that is as capable as its older midsize Llama 4 variant for an order of magnitude less compute.

That efficiency number is worth noting. At the scale Meta operates, compute costs compound fast, and running a frontier-class Meta AI model at a fraction of the cost of its predecessors changes the economics of deploying it in billions of interactions daily.

On benchmarks, the picture is genuinely mixed. Muse Spark scores 52 on the Artificial Intelligence Index v4.0, placing it fourth overall behind Gemini 3.1 Pro, GPT-5.4, and Claude Opus 4.6. Meta has not claimed to have built the best model in the world, which is itself a departure from the over-claiming that damaged Llama 4’s credibility.

Where Muse Spark leads is health. On HealthBench Hard – open-ended health queries – it scores 42.8, substantially ahead of Gemini 3.1 Pro at 20.6, GPT-5.4 at 40.1, and Grok 4.2 at 20.3. Health is a stated priority for Meta; the company says it worked with over 1,000 physicians to curate training data for the model.

Muse Spark also offers three modes of interaction: Instant mode for quick answers, Thinking mode for multi-step reasoning tasks, and Contemplating mode, which orchestrates multiple agents’ reasoning in parallel to compete with the most demanding reasoning modes from Gemini Deep Think and GPT Pro.

The open-source retreat

This is the part of the Muse Spark story that the benchmark tables do not capture. Unlike Meta’s previous models, which were released as open-weight models – meaning anyone could download and run them on their own equipment – Muse Spark is entirely proprietary. The company said it will offer the model in a private preview to select partners through an API, making Muse Spark even more proprietary than the paid models offered by Meta’s rivals.

Wang addressed the change directly, stating: “Nine months ago, we rebuilt our AI stack from scratch. New infrastructure, new architecture, new data pipelines. This is step one. Bigger models are already in development with plans to open-source future versions.”

The developer community’s response has been sceptical. Some see this as a necessary pivot after Llama 4 failed to gain expected traction. Others view it as Meta closing the gates once it has something worth protecting. That is the community now being asked to wait while competitors without that open-source legacy continue shipping freely available weights.

Distribution over benchmarks

Meanwhile, Meta is not waiting for the developer community to come around. Muse Spark will debut in the coming weeks inside Facebook, Instagram, WhatsApp, and Messenger, as well as in Meta’s Ray-Ban AI glasses. That rollout path is arguably more consequential than any benchmark result. OpenAI and Anthropic sell to developers and enterprises. Meta deploys directly to over three billion people already inside its apps daily.

Meta’s push into health does raise privacy questions worth watching. Muse Spark users will need to log in with an existing Meta account to use it, and while Meta does not explicitly say personal account information will be used by the AI, the company has generally trained on public user data and has positioned Muse Spark as a personal superintelligence product.

Meta stock rose more than 9% on the day of the launch, a signal that investors read the Muse Spark release as proof that the US$14.3 billion bet on Wang and the nine-month rebuild produced something real. Whether the promised open-source versions actually materialise is a question the developer community will press every quarter. The answer will define how this chapter of Meta’s AI story is remembered.

See Also: The Meta-Manus review: What enterprise AI buyers need to know about cross-border compliance risk

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Meta has a competitive AI model but loses its open-source identity appeared first on AI News.

Anthropic’s most capable AI model has already found thousands of AI cybersecurity vulnerabilities across every major operating system and web browser. The company’s response was not to release it, but to quietly hand it to the organisations responsible for keeping the internet running.

That model is Claude Mythos Preview, and the initiative is called Project Glasswing.

The launch partners include Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, Nvidia, and Palo Alto Networks. 

Beyond that core group, Anthropic has extended access to over 40 additional organisations that build or maintain critical software infrastructure. Anthropic is committing up to US$100 million in usage credits for Mythos Preview across the effort, along with US$4 million in direct donations to open-source security organisations. 

A model that outgrew its own benchmarks

Mythos Preview was not specifically trained for cybersecurity work. Anthropic said the capabilities “emerged as a downstream consequence of general improvements in code, reasoning, and autonomy”, and that the same improvements making the model better at patching vulnerabilities also make it better at exploiting them. 

That last part matters. Mythos Preview has improved to the extent that it mostly saturates existing security benchmarks, forcing Anthropic to shift its focus to novel real-world tasks–specifically, zero-day vulnerabilities. These flaws were previously unknown to the software’s developers. 

Among the findings: a 27-year-old bug in OpenBSD, an operating system known for its strong security posture. In another case, the model fully autonomously identified and exploited a 17-year-old remote code execution vulnerability in FreeBSD–CVE-2026-4747–that allows an unauthenticated user anywhere on the internet to obtain complete control of a server running NFS. No human was involved in the discovery or exploitation after the initial prompt to find the bug. 

Nicholas Carlini from Anthropic’s research team described the model’s ability to chain together vulnerabilities: “This model can create exploits out of three, four, or sometimes five vulnerabilities that in sequence give you some kind of very sophisticated end outcome. I’ve found more bugs in the last couple of weeks than I found in the rest of my life combined.” 

Why is it not being released?

“We do not plan to make Claude Mythos Preview generally available due to its cybersecurity capabilities,” Newton Cheng, Frontier Red Team Cyber Lead at Anthropic, said. “Given the rate of AI progress, it will not be long before such capabilities proliferate, potentially beyond actors who are committed to deploying them safely. The fallout–for economies, public safety, and national security–could be severe.” 

This is not hypothetical. Anthropic had previously disclosed what it described as the first documented case of a cyberattack largely executed by AI–a Chinese state-sponsored group that used AI agents to autonomously infiltrate roughly 30 global targets, with AI handling the majority of tactical operations independently. 

The company has also privately briefed senior US government officials on Mythos Preview’s full capabilities. The intelligence community is now actively weighing how the model could reshape both offensive and defensive hacking operations. 

The open-source problem

One dimension of Project Glasswing that goes beyond the headline coalition: open-source software. Jim Zemlin, CEO of the Linux Foundation, put it plainly: “In the past, security expertise has been a luxury reserved for organisations with large security teams. Open-source maintainers, whose software underpins much of the world’s critical infrastructure, have historically been left to figure out security on their own.”

Anthropic has donated US$2.5 million to Alpha-Omega and OpenSSF through the Linux Foundation, and US$1.5 million to the Apache Software Foundation–giving maintainers of critical open-source codebases access to AI cybersecurity vulnerability scanning at a scale that was previously out of reach.

What comes next

Anthropic says its eventual goal is to deploy Mythos-class models at scale, but only when new safeguards are in place. The company plans to launch new safeguards with an upcoming Claude Opus model first, allowing it to refine them with a model that does not pose the same level of risk as Mythos Preview. 

The competitive picture is already shifting around it. When OpenAI released GPT-5.3-Codex in February, the company called it the first model it had classified as high-capability for cybersecurity tasks under its Preparedness Framework. Anthropic’s move with Glasswing signals that the frontier labs see controlled deployment–not open release–as the emerging standard for models at this capability level.

Whether that standard holds as these capabilities spread further is, at this point, an open question that no single initiative can answer.

See Also: Anthropic’s refusal to arm AI is exactly why the UK wants it

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Anthropic keeps new AI model private after it finds thousands of external vulnerabilities appeared first on AI News.

The failure mode for enterprise AI in 2026 is not what most people expected. It is not that the models are wrong, or that agents cannot reason, or that the technology is overhyped. The failure mode is that the data feeding those systems is fragmented, inconsistently labelled, and spread across dozens of applications that were never designed to share context. 

Boomi calls this the agentic AI data activation problem, and after tracking 75,000 AI agents running in production across its customer base, the company says solving it comes before everything else. That figure comes from February, when Boomi reported its strongest momentum to date: more than 30,000 customers globally, 75,000 AI agents in production, and a customer base that includes over a quarter of the Fortune 500. 

Yet the consistent pattern across those deployments, according to Steve Lucas, chairman and CEO of Boomi, is that AI value only materialises once the data problem is resolved. “AI only delivers value when data is properly activated, trusted and governed first,” Lucas said when the company announced its latest platform capabilities on March 9.

The fragmentation problem

Enterprise data is not missing; it exists in abundance, distributed across ERP systems, CRMs, data lakes, SaaS platforms, and legacy applications that have accumulated over decades. What is missing is the shared context that allows an AI agent to treat data from one system as reliably compatible with data from another. 

An agent drawing customer records from a CRM and pricing data from an ERP may be working from conflicting definitions of what a customer or a product actually is. The outputs it produces are only as coherent as the data standards beneath them.

Boomi’s answer is Meta Hub, a central system of record announced in its March 9 platform update, designed to standardise business definitions across the enterprise and extend that context to every AI agent operating within it. The goal is to ensure agents reason from a consistent understanding of business logic rather than generating outputs based on fragmented interpretations pulled from disconnected systems.

The same release introduced real-time SAP data extraction via change data capture, addressing one of the most common integration bottlenecks in large enterprises, where SAP data is often inaccessible due to slow, manual export processes that render it effectively unavailable to AI workflows in real-time. 

New governance capabilities for Snowflake Cortex agents within Boomi’s Agent Control Tower added audit trails and session logs, addressing a concern that has moved steadily up enterprise priority lists: AI agents operating as a black box, taking actions with no visible reasoning chain.

What the analyst’s recognition signals

Two independent assessments in March gave Boomi external validation of its positioning. On March 16, Gartner named Boomi a Leader in its 2026 Magic Quadrant for Integration Platform as a Service–the twelfth consecutive time–and positioned it highest for Ability to Execute. 

On March 31, the IDC MarketScape for Worldwide API Management named Boomi a Leader, specifically noting its AI-centric strategy that treats APIs as both the fuel and the control plane for AI workloads. The Gartner framing is pointed. 

The report stated that AI-ready integration is a strategic capability that aligns architecture, integration, and governance to enable AI agents to effectively access enterprise data and operate within business processes. That framing validates the problem Boomi is addressing and signals that iPaaS platforms are now being evaluated on AI readiness rather than traditional integration capabilities alone.

The broader pattern

By now, we are aware that the shift from pilot to production in enterprise AI is stalling in a predictable place. Organisations have models. They have agents. What many do not have is the data infrastructure that makes those agents reliable enough to trust with real business processes.

Data activation–moving data from static storage into live, governed, context-rich flows that agents can actually reason from–is one articulation of what that missing layer needs to look like. Whether that framing becomes the industry standard or gets absorbed into a broader category is a question 2026 will start to answer. 

What is not in question is that the enterprises finding ROI from agentic AI are the ones that sorted the data layer first.

Boomi will be exhibiting at the AI & Big Data Expo at TechEx North America, taking place 18–19 May 2026 at the San Jose McEnery Convention Centre.

(Photo by Boomi)

See also: Autonomous AI systems depend on data governance

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Boomi calls it “data activation” and says it’s the missing step in every AI deployment appeared first on AI News.

The Anthropic UK expansion story is less about diplomatic courtship and more about what happens when a government punishes a company for having principles. In late February, US Defence Secretary Pete Hegseth gave Anthropic CEO Dario Amodei a stark ultimatum: remove guardrails preventing Claude from being used for fully autonomous weapons and domestic mass surveillance, or face consequences. 

Amodei didn’t budge. He wrote that Anthropic could not “in good conscience” grant the Pentagon’s request, arguing that some uses of AI “can undermine rather than defend democratic values.” Washington’s response was swift. 

Trump directed every federal agency to immediately cease all use of Anthropic’s technology, and the Pentagon designated the company a supply chain risk, a label ordinarily reserved for adversarial foreign entities like Huawei. The US$200 million Pentagon contract was pulled. 

Defence tech companies instructed employees to stop using Claude and switch to alternatives. London, watching all of this unfold, saw something different.

The UK’s pitch

Staff at the UK’s Department for Science, Innovation and Technology (DSIT) have drawn up proposals for the US$380 billion company, ranging from a dual stock listing on the London Stock Exchange to an office expansion in the capital, according to multiple people with knowledge of the plans. Prime Minister Keir Starmer’s office has backed the effort, which will be put to Amodei when he visits in late May. 

Anthropic already has around 200 employees in Britain and appointed former prime minister Rishi Sunak as a senior adviser last year. The infrastructure for a meaningful UK presence is already there. What the British government is now offering is an explicit signal that Anthropic’s approach to AI–built on embedded ethical constraints–is an asset, not an obstacle.

A dual listing in London, if it materialised, would give Anthropic access to European institutional investors at a moment when its domestic regulatory standing remains under active legal challenge. The Pentagon’s appeal of the court-ordered injunction blocking the supply chain designation is still before the Ninth Circuit, and the outcome remains uncertain.

Ethics as a competitive advantage

The dispute has been framed largely as a legal and political fight. But its implications for global AI governance run deeper. Anthropic’s lawyers argued in court filings that Claude was not developed to be used for lethal autonomous weapons without human oversight, nor deployed to spy on US citizens, and that using the tools in these ways would represent an abuse of its technology. 

US District Judge Rita Lin, who granted a preliminary injunction blocking the blacklist in March, found the government’s actions “troubling” and concluded they likely violated the law. That judicial finding matters in the UK context. Britain is positioning itself as a regulatory environment sitting between Washington’s current posture, which demands unrestricted military access, and Brussels, where the EU AI Act imposes its own constraints. 

The UK government presents itself as offering a less constrained environment for AI companies than either the US or the European Union. Crucially, that pitch doesn’t ask Anthropic to abandon the guardrails it went to court to defend.

The courtship also sits alongside broader UK efforts to build domestic AI capability, including a recently announced £40 million state-backed research lab, after officials acknowledged the absence of a homegrown competitor to the leading US frontier labs.

Competition in London

The UK’s play for Anthropic is not happening in a vacuum. OpenAI has already committed to making London its biggest research hub outside the US. Google has anchored itself in King’s Cross since acquiring DeepMind in 2014. The race to secure frontier AI in London is already competitive, and Anthropic’s current circumstances make it the most consequential target yet.

Anthropic has been expanding internationally regardless of its domestic legal battles, including opening a Sydney office as its fourth Asia-Pacific location. The global growth strategy is already in motion. What remains to be seen is how much of it London gets to claim.

The company Washington blacklisted for having an AI ethics policy is now being actively courted by another G7 government that wants exactly that. The late May meetings with Amodei will be telling.

See Also: Anthropic selected to build government AI assistant pilot

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Anthropic’s refusal to arm AI is exactly why the UK wants it appeared first on AI News.