Tech

China’s Fengwu AI Forecasted Typhoon Landfall Five Days Ahead

Published

39 minutes ago

August 12, 2026

A typhoon gave China’s AI weather models a real-world test — and one system reportedly narrowed down the storm’s landfall days in advance.

Chinese researchers are using AI models to respond to the country’s current weather conditions proactively.

According to Reuters, Shanghai AI Lab’s Fengwu predicted the typhoon’s landfall location and timing five days ahead to within about 30 kilometers and 30 minutes. The result highlights how AI-based forecasting systems could complement conventional weather models by producing useful predictions quickly and with less computing power.

Fengwu is one of several AI weather systems being developed in China, alongside Huawei’s Pangu and Fudan University’s Fuxi, as meteorologists test whether machine-learning models can improve forecasting without replacing traditional physics-based systems or human expertise.

According to The Independent, its developers reported it outperformed Google’s GraphCast on 80% of evaluated weather variables in tests, extending skillful global medium-range forecasting beyond 10 days. That does not mean Fengwu is 80% more accurate overall. The figure refers specifically to the proportion of evaluated variables where it performed better.

Huawei has developed Pangu, while Fudan University has developed Fuxi, giving Chinese meteorologists multiple AI-based systems for forecasting atmospheric conditions.

The models work differently from traditional numerical weather prediction. Conventional systems use supercomputers to solve mathematical equations that represent atmospheric physics. AI models, on the other hand, are trained on vast datasets of historical weather data, enabling them to learn patterns and predict how atmospheric conditions will change.

Once trained, these models can produce forecasts with far less computation.

Why AI weather models are getting attention

The bigger story is not that AI can forecast weather; it is that prediction itself is becoming one of AI’s most practical uses.

The same basic approach can also be applied to problems such as wildfire risk and industrial equipment failures, where identifying a likely future event early can be more valuable than simply describing what is happening now.

Beyond these, there’s also the use of AI in prediction markets like sports and stocks.

Taken together, these give predictive AI a different role from the generative systems that have dominated public attention. Instead of asking an AI to create something, the user is effectively asking it to assess the available evidence and estimate the future.

AI systems aren’t replacing human experts

AI systems can be very good at learning atmospheric patterns from historical data, but weather is a chaotic physical system. Some events are difficult to predict because small differences in atmospheric conditions can produce markedly different outcomes, making it hard for AI models to forecast every weather event accurately.

Besides, AI systems tend to have certain weaknesses, particularly when dealing with edge cases, which are very common in weather forecasting. An unusual combination of weather conditions or a rapidly developing storm can create situations the model may not have encountered often enough during training to make an accurate prediction.

That means AI systems are not replacing the human experts who interpret forecasts. These human experts bring the unique human touch with years of experience to assess uncertainty and identify when an AI prediction may not tell the full story.

What this means for everyone

The clearest advantage of AI forecasting may be speed. Once trained, models such as Fengwu can produce forecasts with much less computation than conventional numerical systems.

If those systems can also maintain useful accuracy farther in advance, that could give emergency officials more time to issue warnings, prepare infrastructure and move people out of harm’s way. For now, the likely future is not AI replacing conventional forecasting, but AI becoming another increasingly important tool alongside physics-based models and human meteorologists.

Other News: Japan is planning a new AI data center in Akita, highlighting the push to expand computing infrastructure beyond the country’s traditional technology hubs.

Tech

Report: iOS 27 Beta Reveals Six Unreleased iPhone Codenames

Published

4 minutes ago

August 12, 2026

qasimgillani

Apple’s iOS 27 beta 5 reportedly contains internal references to six unreleased iPhone models, offering an unusually broad glimpse at hardware the company has yet to announce. Macworld first reported that the identifiers appear in battery-related system files alongside references to existing iPhones.

The codes align with previous reporting that has linked them to the iPhone 18 family, a second-generation iPhone Air, and Apple’s expected foldable iPhone. They also fit reports that Apple may be preparing a wider lineup with launches spread across fall 2026 and spring 2027. The beta does not reveal final product names, specifications, prices, or release dates, and Apple has not announced any of the devices.

Six codes map Apple’s expanding iPhone lineup

The identifiers are said to be V62, V63, V64, V67, V68, and V69. The Information’s reporting on Apple’s iPhone roadmap previously linked them to these planned devices:

V62: second-generation iPhone Air
V63: iPhone 18 Pro
V64: iPhone 18 Pro Max
V67: iPhone 18
V68: foldable iPhone
V69: iPhone 18e

Macworld’s analysis of the beta found the identifiers in Battery Intelligence and battery-driver files. Their presence suggests the projects remain represented in Apple’s software development, although internal references do not establish whether every device will ultimately ship.

V68’s retail name remains uncertain. Reports use both “iPhone Fold” and “iPhone Ultra,” but Apple has confirmed neither. TechRepublic has been tracking Apple’s expected foldable iPhone as multiple reports have pointed to a 2026 launch window for the company’s first folding handset.

The foldable could also broaden the range of screen sizes and form factors that developers and enterprise IT teams support within Apple’s mobile ecosystem. Applications that behave differently across screen configurations or multitasking modes could require additional validation once Apple publishes specifications and developer guidance.

The codes seem to reinforce a changing release strategy

Battery-related software offered an earlier clue about Apple’s foldable plans. On July 20, iOS 27 beta 4 introduced system strings referring to multiple batteries in an iPhone. Foldable phones commonly divide battery capacity between sections of the device, although Apple has not confirmed that the strings refer to its unannounced handset. Earlier TechRepublic coverage also examined foldable-related clues found in iOS 27.

The six identifiers also align with reports that Apple may stagger its next iPhone release cycle. Current reporting points to a fall 2026 launch for the iPhone 18 Pro, iPhone 18 Pro Max, and foldable, with the standard iPhone 18, iPhone 18e, and redesigned second-generation iPhone Air reportedly targeted for spring 2027. Apple has not confirmed those launch windows.

TechRepublic previously reported that the standard iPhone 18 could shift to 2027, a change that would break from Apple’s usual practice of concentrating major numbered iPhone launches in the fall.

A split rollout could spread device testing, mobile device management validation, app compatibility checks, and procurement across separate release windows. Organizations that standardize employee iPhones could also face different evaluation and purchasing cycles for mainstream and premium hardware.

Until Apple announces the devices and publishes compatibility requirements, the six identifiers remain an early view of its development roadmap rather than a firm fleet-planning timetable.

Read more: Apple’s lineup could expand further after this cycle, with reports pointing to an even broader iPhone family in 2027 as the company reshapes how and when it introduces new models.

Tech

After Microsoft threatened legal action, a security researcher publishes a new Windows zero-day bug

Published

11 minutes ago

August 12, 2026

qasimgillani

A security researcher has published details of a new vulnerability in the latest versions of Windows that allows hackers to gain system-wide access to the user’s device and data, despite facing a legal threat from Microsoft weeks earlier over the release of previously unknown software flaws.

The new bug, dubbed ShieldBreak, is the latest disclosure by security researcher Nightmare Eclipse, who in recent months has published details of several bugs affecting Microsoft’s products, including Windows.

According to Nightmare Eclipse’s post, ShieldBreak takes advantage of a flaw in Windows Defender, the anti-malware and security engine built into Windows. A successful attack allows the hacker to escalate their permissions from a low-level user to full access to the device and its data.

Nightmare Eclipse published the proof-of-concept exploit as a Windows app, requiring the user to run the app to exploit the bug. The bug works on Windows 10, Windows 11 (including the latest 25H2 version), and Windows Server 2025, the researcher said.

Security researcher Will Dormann verified that the bug works and that Windows Defender must be enabled for the exploit to work.

The latest exploit builds on an earlier exploit that Nightmare Eclipse developed dubbed RoguePlanet. Microsoft rolled out a patch for RoguePlanet, but Nightmare Eclipse implied that Microsoft’s fix was not sufficient and that their latest exploit demonstrates a full bypass of the earlier patch.

Microsoft has not yet released a patch for the ShieldBreak bug. A spokesperson for Microsoft did not immediately comment when contacted by TechCrunch. The bug is considered a zero-day because the software maker — in this case, Microsoft — was given no time to patch the bug before it was publicly disclosed.

The release of this new zero-day is the latest in a long back-and-forth between the security researcher and the software giant over the company’s alleged handling of their bug reports.

In a series of blog posts, the security researcher claimed that Microsoft mistreated them and did not handle their bug reports sufficiently, with the implication that the researcher had no other choice but to publicly disclose the bugs online. Nightmare Eclipse previously released several other bugs in Windows that were later exploited in real-world attacks to hack into organizations.

In May, Microsoft published a blog post threatening to take legal action against security researchers, like Nightmare Eclipse, if they released details of zero-days outside of the company’s disclosure policies. The company faced heavy rebuke from the security community, many of whom described similar experiences with Microsoft’s handling of their bug reports. Microsoft later walked back the comments in a social media post. Its original blog post remains published and unchanged.

ShieldBreak lands a day after Microsoft’s regularly scheduled monthly security patch releases, dubbed Patch Tuesday. This is the second month in a row where the number of patches has reached around 500 or so bugs driven by the company’s growing use of AI to find and weed out security flaws.

When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.

Tech

Could AI Increase Fossil Fuel Emissions in APAC Oil and Gas?

Published

18 minutes ago

August 12, 2026

qasimgillani

AI could add to fossil fuel emissions in a less obvious way than power-hungry data centers: by making oil and gas production cheaper and more productive. New research suggests those gains could outweigh some of AI’s benefits for cleaner energy.

The peer-reviewed research in npj Climate Action modeled AI-driven productivity gains across fossil fuels and renewable energy. Across 64 scenarios, researchers estimated a net annual increase of 0.47 billion to 1.8 billion metric tons of carbon dioxide when AI improved both sectors. The results represent modeled economic effects, not measured emissions or forecasts for individual companies or countries.

The study examines what its authors call “enabled emissions”: additional emissions that can result when AI lowers costs or raises productivity in fossil fuel extraction, processing, and energy production. APAC energy companies are already expanding AI across upstream operations.

Malaysia’s PETRONAS Carigali said July 8 that it was expanding its TriCipta AI initiative with IBM and Tridiagonal.AI. The work targets surface-equipment optimization and production and maintenance decisions, while earlier tools have supported geoscience and exploration analysis.

AI efficiency could drive more fossil fuel production

Lower operating costs do not automatically mean lower emissions. AI that reduces exploration costs, improves recovery rates, or makes existing assets cheaper to operate could make additional fossil fuel production economically viable.

Australia’s Woodside provides another example of upstream AI adoption. Its Maint Intel system analyzes maintenance records and equipment performance to recommend maintenance intervals at the North West Shelf project. Woodside said testing on the offshore Angel platform cut model-processing time from five days to under two hours.

Neither deployment shows that AI has increased emissions at PETRONAS or Woodside. Both demonstrate the kinds of operational productivity gains examined by the global research. AI can also support methane detection, equipment reliability, and other emissions-reduction efforts.

AI’s electricity use creates a separate emissions footprint. The International Energy Agency expects global data-center electricity consumption to roughly double from 485 TWh in 2025 to 950 TWh in 2030. TechRepublic has separately covered grid pressure in Australia and changing power and cooling requirements for AI infrastructure. A planned 360 MW Nvidia-powered AI data center in Indonesia shows how quickly regional capacity is growing.

APAC methane cuts lag technical potential

The IEA estimates fossil fuel operations in South and Southeast Asia emitted about 13 million metric tons of methane in 2025. More than 60% came from coal, with the remainder from oil and gas. India and Indonesia were the region’s largest fossil fuel methane emitters.

Existing technology could cut methane emissions across South and Southeast Asia by more than 50%, with 60% of those reductions achievable at no net cost to producers. Under stated policies, emissions are projected to fall only 10% by 2030 and almost 20% by 2035.

China faces a similar gap. More than 90% of available methane reductions in its oil and gas sector could be achieved at no net cost, according to the IEA.

Greater operating efficiency does not necessarily reduce absolute emissions. Operators procuring AI tools should track production KPIs and emissions KPIs separately so efficiency gains are not treated as evidence of a climate benefit without a measured reduction in emissions.

Read more: Australia’s AI boom is also reshaping capital spending, with data centers accounting for a growing share of private investment as demand for compute infrastructure accelerates.