A massive magnet that crossed the Atlantic Ocean has brought one of America’s most ambitious fusion projects a step closer to reality. The U.S. Department of Energy’s Princeton Plasma Physics Laboratory recently received the central magnet bundle for its upgraded spherical tokamak, a compact fusion system designed to test technologies that could shape future power plants.

The delivery marks a critical milestone for the machine’s return to operation. Engineers now enter the final stages of assembly before experiments resume in 2027.

Manufactured by Spain-based Elytt Energy, the component traveled from Bilbao to New Jersey aboard a cargo aircraft before making its final trip by flatbed truck to Princeton. Weighing 23,000 pounds and stretching roughly 20 feet long, the bundle ranks among the most important pieces of the fusion device.

Dave Micheletti, associate laboratory director for engineering and project director, called the arrival “a truly momentous occasion.” He said the team will now focus on completing assembly and preparing the system for scientific use.

Magnetic core

The central bundle combines two critical magnet systems into a single structure. The toroidal field magnets generate the primary magnetic fields that confine the plasma. Meanwhile, the ohmic-heating system drives electrical current through the plasma to raise temperatures and support confinement.

Building the assembly required a complex manufacturing process. Technicians joined 36 copper conductors, each measuring 19 feet long, using fiberglass tape, resin, and vacuum-pressure impregnation techniques.

Engineers then wrapped additional copper coils around the structure to create the ohmic-heating system. They sealed the completed assembly with insulating resin to withstand demanding operating conditions.

Once installed, the magnets will create two magnetic field configurations inside the apple-shaped vacuum vessel. Together, they stabilize and heat the plasma needed for fusion experiments.

Testing future reactors

Researchers believe spherical tokamaks could offer advantages over conventional designs. Their compact shape may improve efficiency while reducing construction and replication costs. The upgraded system will help scientists evaluate whether the approach can support future commercial fusion plants.

The machine also supports broader national goals. Researchers plan to gather operational data that could train artificial intelligence tools capable of improving fusion performance and control strategies.

Jonathan Menard, the laboratory’s deputy director for research, said the facility will provide new opportunities for scientists and engineers entering the field. He encouraged researchers worldwide to use the platform to advance fusion science.

Final assembly begins

Engineers have already positioned the magnet bundle near the fusion device using heavy-duty cranes. Over the coming months, teams will rotate the structure upright and prepare it for installation. Workers will lower protective shielding around the assembly before placing it inside the machine’s interior.

Technicians must still connect dozens of flexbus components, cooling lines, and internal systems. They will also complete bakeout procedures and commissioning tests to verify that all systems operate together safely.

Steven Cowley, the laboratory’s director, said the upgraded machine possesses capabilities unavailable elsewhere. He noted that the project could play an important role in determining the future direction of commercial fusion research.

If the remaining work proceeds as planned, the upgraded facility could begin experiments in 2027, giving American researchers a powerful new platform to explore the promise of fusion energy.

💾

Metal additive manufacturing continues to expand beyond aerospace prototypes and into production-focused energy systems. That shift gained momentum this week as Velo3D and gas turbine developer Aurelia Technologies unveiled a partnership aimed at bringing 3D printing deeper into next-generation turbine development.

The companies plan to evaluate, qualify, and eventually manufacture turbine components using additive techniques. Their goal extends beyond experimentation. They want to simplify designs, shorten development cycles, and build a more resilient supply chain for future turbine programs.

The collaboration reflects a broader trend across U.S. manufacturing. Engineers increasingly use additive manufacturing to solve practical production challenges rather than showcase technological novelty. Faster iteration, fewer components, and lower inventory exposure now drive many adoption decisions.

Designing out complexity

Aurelia develops small-scale gas turbines designed to operate efficiently while accommodating multiple fuel types. As its turbine platforms evolve, the company sees additive manufacturing as a way to remove unnecessary complexity from critical assemblies.

Instead of assembling numerous individual pieces, engineers can consolidate multiple components into a single printed part. That approach reduces fasteners, minimizes interfaces, and lowers the risk associated with tolerance stack-ups.

The result could improve durability in demanding environments where turbines face extreme temperatures and mechanical stress.

“Additive manufacturing allows us to simplify designs, reduce failure points, and move faster,” said Karol Hricisak, director of technology at Aurelia Technologies. He added that the company remains grounded in established turbomachinery principles and proven materials science.

Reducing part counts also supports maintenance objectives. Fewer connections can translate into fewer potential failure points throughout a turbine’s operating life.

Testing to production

The partnership will unfold through several phases. Early efforts will focus on determining which turbine components benefit most from additive production.

The companies also plan to assess suitable alloys and refine manufacturing processes before advancing toward qualification activities. The long-term objective involves low-rate initial production using Velo3D’s Sapphire XC metal additive manufacturing platform.

Engineers will examine whether additive methods can improve manufacturability while reducing lead times for complex parts.

Unlike conventional production methods, additive manufacturing can accommodate geometry changes without extensive tooling modifications. Design updates that traditionally require months could move into production within weeks.

Strengthening domestic supply chains

The initiative also addresses supply chain concerns that continue to affect advanced manufacturing sectors across the United States.

Traditional turbine production often depends on long-lead forgings, specialized tooling, and significant inventory commitments. Additive manufacturing offers an alternative path that can reduce those dependencies.

By producing components closer to demand, manufacturers can respond more quickly to engineering changes and shifting market requirements.

Velo3D said Aurelia selected the company because of its experience in metal additive manufacturing and its structured approach to qualification and production scaling.

“Advanced energy systems are pushing the limits of traditional manufacturing,” said Michelle Sidwell, Velo3D’s chief revenue officer. She said additive manufacturing delivers the greatest value when engineers incorporate it early in the design process.

As the initial work progresses, both companies will evaluate additional applications and opportunities to expand the collaboration. The effort highlights how additive manufacturing increasingly serves as a practical engineering tool for next-generation energy and propulsion systems.

Otto Aerospace has cleared a major regulatory hurdle as it pushes its unconventional Phantom 3500 business jet closer to certification. The milestone gives the Texas aircraft maker a defined path forward as it develops one of the industry’s most closely watched clean-sheet aircraft programs.

The Federal Aviation Administration has finalized the Phantom 3500’s G-1 Issue Paper, establishing the aircraft’s certification basis under 14 CFR Part 23. The move allows Otto to transition from defining requirements to proving that its ambitious design can meet them. The company continues to target first flight in 2027 and entry into service by 2030.

Defining the path ahead

Otto became an FAA applicant for type certification in September 2025 and chose to pursue approval under Part 23. The company sought to take advantage of efficiencies introduced through Amendment 23-64, which streamlined parts of the certification process.

Closing the G-1 Issue Paper defines the airworthiness standards the Phantom 3500 must satisfy before entering service. Otto is also working with the FAA’s East Certification Branch to complete the G-2 Issue Paper, which outlines how compliance will be demonstrated throughout the certification campaign.

Scott Drennan, president and chief executive officer of Otto Aerospace, called the milestone a significant moment for the program. He said the agreement with regulators provides clarity on what the company must demonstrate and gives the team momentum as it moves toward first flight.

Otto recently completed its Preliminary Design Review and has started advanced material testing to support the next stages of development.

Rethinking jet efficiency

The Phantom 3500 has drawn industry attention because of its unconventional approach to aircraft design. Otto’s strategy centers on transonic laminar flow technology, which aims to keep airflow smooth across the aircraft’s surface and reduce the turbulence that creates drag.

The company recently completed an unmanned flight-test campaign in New Mexico to validate the aerodynamic principles behind the concept. Those same principles also explain the Phantom 3500’s windowless cabin design.

Traditional passenger windows interrupt airflow and make it harder to maintain laminar flow along the fuselage. Otto instead plans to use digital displays connected to external cameras, creating a cleaner exterior profile.

The company says combining laminar-flow aerodynamics with carbon-fiber composites could reduce fuel consumption by 61% compared with today’s super-midsize business jets. Otto also claims the design could deliver 35% less drag and a striking 90% reduction in emissions.

Toward first flight

Otto plans to conduct flight testing from Cecil Airport in Jacksonville, Florida, where the company has established its new operational base. Engineers continue advancing certification work while preparing for the first flight of the Phantom 3500’s test vehicle in 2027.

Kerri Hinton, Otto’s director of test and certification, said safety remains central to the program as the company works alongside regulators to move the aircraft through the approval process.

The Phantom 3500 also enters its next phase with substantial commercial backing. Flexjet has committed to purchasing 300 aircraft in a deal worth more than $5 billion, marking one of the largest orders in business aviation history.

For Otto Aerospace, regulatory progress now shifts the focus to execution. The company must prove that its bold engineering ideas can perform in real-world conditions and redefine expectations for business jet travel.

Anthropic has opened the doors to its most advanced AI system yet, but not without placing guardrails around its most sensitive capabilities.

The company on Tuesday unveiled Claude Fable 5, the first publicly accessible version of its Mythos-class AI model. Anthropic says the model outperforms every generally available system it has released to date, delivering major gains in software development, analytical work, visual understanding, and scientific reasoning.

The launch marks a significant shift for the AI firm. Until now, Mythos-level capabilities remained restricted to a small group of trusted organizations because of concerns that the technology could aid cyberattacks and other harmful activities. Fable 5 changes that equation by bringing those capabilities to a broader audience while routing certain high-risk requests through a less permissive model.

Safety before scale

Anthropic said Fable 5 includes a new layer of AI-powered classifiers designed to detect potentially dangerous prompts and jailbreak attempts. The company stress-tested those safeguards before releasing the model to the public.

If users submit requests involving cybersecurity, biology, chemistry, or model distillation, the system automatically redirects those interactions to Claude Opus 4.8 instead of Fable 5. Users receive a notification whenever that fallback occurs, allowing them to understand why the switch happened.

Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use.

Its capabilities exceed those of any model we’ve ever made generally available. pic.twitter.com/2AvmEjHIX8

— Claude (@claudeai) June 9, 2026

The company acknowledged that the safeguards may occasionally block harmless requests. However, Anthropic said early testing showed that fewer than 5% of sessions triggered a fallback response. “We’ve deliberately tuned the safeguards to be cautious,” the company said, adding that it plans to refine the system over time to reduce unnecessary interventions.

Built for demanding work

Anthropic argues that Fable 5 delivers its biggest advantages when tackling lengthy and complex assignments. The company says the model can stay focused across extended workflows while maintaining consistency over large amounts of context.

During early evaluations, payments company Stripe reported that the model completed a large-scale migration across a 50-million-line Ruby codebase in roughly one day. According to Anthropic, the same task would have otherwise occupied an engineering team for more than two months.

The company also highlighted gains in financial analysis. On Hebbia’s finance benchmark, Fable 5 posted the strongest results among leading models. Trading firm IMC found that the system performed well across factual analysis, reasoning tasks, root-cause investigations, and expected-value calculations.

Fable 5 also advanced Anthropic’s vision capabilities. The model can extract detailed information from scientific figures and reconstruct web applications from screenshots. In another demonstration, it completed Pokémon FireRed using a minimal vision-only setup after earlier Claude models struggled with the challenge.

Mythos stays restricted

Alongside Fable 5, Anthropic introduced Claude Mythos 5, a version of the same underlying model with fewer restrictions in select areas. Anthropic says the model offers the strongest cybersecurity capabilities it has developed so far.

Access to Mythos 5 remains limited. The company is deploying it through Project Glasswing while expanding availability to approved cyber defenders and organizations responsible for critical infrastructure. Anthropic said it intends to broaden access further through a trusted program in the future.

Fable 5 is available through Anthropic’s Claude API and consumption-based Enterprise offerings. Through June 22, Pro, Max, Team, and seat-based Enterprise subscribers can access the model at no additional cost. After that date, Anthropic plans to move users to a credit-based system before eventually restoring broader subscription access.

As competition intensifies across the AI industry, Anthropic is betting that powerful models do not have to arrive without safeguards. With Fable 5, the company is testing whether frontier capabilities and tighter controls can coexist in the hands of everyday users.

💾

The next frontier of AI involves machines that can move, see, and make decisions in the physical world. NVIDIA and South Korea’s LG Group are betting heavily on that future through a broad collaboration aimed at weaving AI into everything from factory floors and household robots to autonomous vehicles and large-scale data centers.

The companies unveiled plans to develop a connected AI ecosystem that links model development, simulation, deployment, and infrastructure under a single framework. By combining NVIDIA’s accelerated computing platforms with LG’s expertise across manufacturing, consumer electronics, telecommunications and mobility, the partners hope to shorten the path from AI experimentation to commercial deployment.

Reinventing factory operations

Manufacturing sits at the center of the partnership. LG plans to use AI to connect every stage of its industrial operations, creating a system that can respond to changing conditions with greater speed and efficiency. The company aims to establish an autonomous manufacturing ecosystem in which procurement, production, logistics, and customer delivery operate through continuous data analysis and real-time decision-making.

To support that vision, LG will combine production expertise gathered from its global facilities with NVIDIA’s digital twin technologies. Engineers will be able to simulate factory environments, test operational changes, and validate processes virtually before implementing them in the real world. The companies believe the approach could help define a new standard for smart manufacturing.

Robots learn virtually

LG is also deepening its push into robotics. The company continues developing home robots designed to assist users with everyday indoor tasks. Instead of relying solely on physical testing, developers will train and evaluate these machines inside realistic virtual environments before deployment.

LG plans to incorporate NVIDIA Isaac Sim and Isaac Lab into its robotics workflows, allowing teams to simulate scenarios and refine robot behavior more efficiently. The company is also exploring NVIDIA’s Isaac GR00T foundation model to equip robots with stronger reasoning capabilities and the ability to handle more sophisticated tasks.

One of robotics’ biggest challenges remains access to quality training data. To address that problem, LG Electronics is building what it describes as a physical AI data factory. Using NVIDIA Cosmos world models, the initiative will generate synthetic datasets to support the development of robotics and industrial AI. LG CNS plans to integrate NVIDIA’s robotics technologies into its industrial automation platform, while LG Innotek intends to develop sensing solutions optimized for NVIDIA systems.

Building AI infrastructure

Beyond robotics, the partnership extends to the computing backbone that supports increasingly demanding AI applications. LG Electronics plans to work with NVIDIA on liquid-cooling technologies and modular infrastructure aligned with the NVIDIA DSX AI factory platform, enabling faster deployment of high-performance computing facilities.

LG Uplus intends to build large-scale AI data centers capable of supporting the latest NVIDIA GPUs and to explore future GPU cloud services. LG Energy Solution is evaluating advanced 800-volt direct-current power technologies designed for next-generation facilities.

The collaboration also reaches the automotive sector. LG Electronics plans to align its advanced driver-assistance technologies with NVIDIA DRIVE platforms to support future software-defined vehicles, AI-powered cockpits and edge computing applications.

Airbus Helicopters has revealed a new autonomous aircraft based on its popular H145 platform, pushing the company deeper into the growing market for large uncrewed rotorcraft. The aircraft, named the U145, debuted as a full-scale mock-up ahead of the ILA Berlin Air Show and signals Airbus’ latest effort to blend proven helicopter performance with advanced autonomous flight systems.

The company plans to fly the aircraft with a safety pilot on board by the end of 2026. Airbus aims to place the platform into operational service early in the next decade.

Unlike many unmanned helicopters designed from scratch, the U145 builds directly on the H145, one of Airbus’ most widely used rotorcraft families. More than 1,800 H145 helicopters currently operate worldwide across military, civil, and public service missions. Together, they have logged over 8.5 million flight hours.

Heavy-lift drone push

Airbus designed the U145 to carry larger payloads and stay airborne longer than many existing rotary-wing drones. The aircraft features a maximum takeoff weight of roughly 8,400 pounds and targets missions ranging from cargo resupply to disaster response.

The company removed the traditional cockpit entirely. Instead, Airbus integrated autonomous systems, artificial intelligence tools, and a dedicated sensor package to manage flight operations without onboard pilots.

The redesign also reshapes the aircraft for logistics work. Engineers added an integrated nose cargo door, a foldable loading table, and a reinforced cargo floor to improve loading efficiency during field operations.

Airbus Helicopters CEO Matthieu Louvot said the U145 combines the H145’s established airframe and lifting power with autonomous operating capabilities usually associated with smaller drones. He added that Airbus plans to work with European autonomy specialists to expand the region’s uncrewed aviation ecosystem.

Military missions expanding

Airbus expects the U145 to support both civilian and military operations. The company described the aircraft as “mission agnostic,” meaning operators can adapt it for multiple roles with minimal structural changes.

Potential missions include firefighting, surveillance, armed reconnaissance, and disaster management. Airbus also wants the aircraft to act as a drone mothership capable of launching smaller airborne systems during combat operations.

The company already partners with missile manufacturer MBDA on air-launched effects technology tied to future European defense programs.

Airbus also sees the aircraft supporting crewed-uncrewed teaming missions, where autonomous aircraft operate alongside traditional helicopters and fighter aircraft.

The project reflects broader defense trends in the United States and Europe. Military planners increasingly want autonomous aircraft to handle dangerous logistics and reconnaissance missions while reducing risks to pilots in contested environments.

Existing platform advantage

The U145 marks Airbus’ second attempt to convert a crewed helicopter into an autonomous aircraft. The company previously developed the VSR700 uncrewed system from the Cabri G2 helicopter platform.

Airbus believes using an already certified helicopter gives the U145 a major advantage over clean-sheet drone designs. Existing supply chains, maintenance networks, and operational data could shorten development timelines and reduce certification risks.

In the United States, Airbus U.S. Space & Defense already offers a related autonomous aircraft concept to the U.S. Marine Corps. The MQ-72C, developed with Shield AI, L3Harris and Parry Labs, derives from the UH-72B Lakota helicopter.

The H145 platform itself also brings operational benefits. Powered by twin Safran Arriel 2E engines with FADEC controls, the helicopter maintains one of the lowest acoustic signatures in its class while producing lower carbon emissions than many competitors.

During his final Worldwide Developers Conference keynote before stepping down as Apple CEO in September, Tim Cook unveiled the long-awaited Siri overhaul that has been at the center of Apple’s AI ambitions for years.

At WWDC 2026, Apple formally introduced Siri AI, a rebuilt version of its digital assistant that arrives after delays and growing pressure from rivals in the artificial intelligence race. The keynote also brought a broad expansion of Apple Intelligence features across the iPhone, iPad, and Mac, making AI the defining theme of Cook’s final developer conference.

Alongside the AI announcements, Apple revealed iOS 27, new family safety tools, refinements to its Liquid Glass design language, and the next version of macOS, called Golden Gate.

Siri’s AI transformation

The headline announcement was Siri AI, a redesigned version of Apple’s voice assistant built to handle more natural conversations and provide more contextual responses. Users can now customize how Siri sounds through new pacing and expressiveness controls, while also interacting with the assistant through voice or text.

Apple has moved Siri into a dedicated app while maintaining integration across its operating systems. The assistant also gains access to Visual Intelligence features, allowing it to identify locations and other details from images stored on a user’s device.

The launch marks a significant milestone for Apple after months of scrutiny surrounding delays to advanced Siri features. The company is positioning the new assistant as a central part of its broader AI strategy moving forward.

Apple Intelligence spreads across iOS

Apple Intelligence is now reaching deeper into many of the company’s most-used applications. Safari can monitor webpages and notify users when information changes, while Messages and Mail gain smarter contextual reply suggestions.

Calendar can create events from conversational prompts, and the Home app can summarize camera activity and make footage easier to search. Apple is also bringing AI-assisted automation to Shortcuts, allowing users to describe a task and have the software build the shortcut automatically.

The updates reflect Apple’s effort to weave AI features directly into everyday workflows rather than confining them to standalone tools.

iOS 27 expands compatibility

Apple also highlighted the scale of its upcoming software rollout. The company said iOS 27 will support every iPhone model dating back to the iPhone 11, making it one of the broadest compatibility expansions in the platform’s history.

Apple claimed the update will reach more users than any previous iOS release. The company also announced several performance improvements, including faster photo loading, quicker AirDrop transfers, and multitasking enhancements.

Alongside iOS 27, Apple introduced new tools for families. Parents will be able to create approved app lists for child accounts and manage access over time. A new Communication Safety feature will help prevent children and teenagers from viewing shared content that contains nudity, graphic imagery, and other sensitive material.

Apple also refined its Liquid Glass interface with new readability improvements and user-controlled transparency settings. The company closed the software showcase with a preview of macOS Golden Gate, which will become the next major release for Mac users.

Taken together, the announcements represented Apple’s most ambitious AI-focused software update yet, giving Cook a fitting final WWDC centered on the Siri upgrade that many users had been waiting to see.

💾

The Pentagon is searching for new ways to protect military satellite communications from electronic attacks in space. MIT Lincoln Laboratory engineers are now developing a lightweight antenna system designed to keep U.S. forces connected even during heavy jamming attempts.

The prototype was built for proliferated low-Earth orbit, or pLEO, where large constellations of smaller satellites require compact and power-efficient hardware. Researchers say the antenna could deliver advanced beamforming capabilities while using far less power than conventional systems.

Lowering satellite power needs

Modern military satellites increasingly rely on adaptive antenna arrays that can rapidly redirect signals and block interference. Those systems help operators maintain links during hostile jamming attempts.

However, traditional phased-array antennas consume large amounts of power and require complex hardware. MIT Lincoln Laboratory’s answer is a scanning reflectarray antenna called Hosted Nimble Beamforming Anti-Jam Reflectarray, or HoNi BAJR.

Instead of using amplifiers across every antenna element, the design relies on a reflective surface with individually controlled elements. Incoming signals bounce off the surface and shift phase before traveling toward a separate feed antenna.

That process shapes and steers the beam while reducing hardware complexity. Engineers estimate the reflectarray design cuts power consumption by roughly 95% compared with traditional array systems.

The smaller footprint also allows the antenna to fit on compact satellite platforms commonly used in pLEO constellations. Researchers believe that advantage could make the technology attractive for future military satellite constellations.

Michael Craton, a technical staff member in the laboratory’s Tactical Satellite Communications Group, said future military satellite networks need scalable systems with lower size, weight, power, and cost requirements.

“We want to think about ways we can achieve exquisite performance using less expensive hardware,” Craton said. He added that the team also wants to prepare for emerging threats before they become operational problems.

Blocking jamming in orbit

The HoNi BAJR system focuses heavily on anti-jamming performance. Adaptive arrays normally counter interference by creating “nulls,” which suppress signals coming from hostile directions.

During testing at the laboratory’s RF Systems Testing Facility, the prototype demonstrated wide-angle scanning capability. Researchers also confirmed the antenna could split beams between multiple users with minimal signal degradation.

The team pushed further by developing broader interference suppression zones rather than targeting single jamming points. Engineers attempted this by reshaping beam side lobes to reduce interference across wider regions.

That approach produced mixed results during early testing because small signal variations affected side-lobe control. Researchers believe improved calibration methods could solve much of the instability.

Calibration remains challenging

Calibration now stands as one of the project’s biggest technical hurdles. Unlike conventional phased arrays, scanning reflectarrays lack extensive operational history in military satellite systems.

Engineers still need to determine how to precisely measure and compensate for signal distortions across the antenna. Researchers say accurate calibration will improve beam steering, interference suppression, and overall system performance.

The laboratory also continues studying where reflectarrays fit best within future military architectures. Early findings suggest the technology works well in scheduled beam operations, power-limited spacecraft, and environments with persistent but less dynamic interference.

“Designing hardware is always a challenge,” Craton said. “But figuring out how to fit the technology into a complete and functional system that meets mission needs is the hardest part.”

Future work will refine beamforming techniques, improve calibration procedures, and explore operational use cases for the antenna technology.