Month: December 2025

China’s fast-advancing artificial intelligence startup DeepSeek has unveiled two upgraded versions of its experimental model, marking a significant step forward in the global race toward next-generation AI systems. The new releases, known as DeepSeek-V3.2 and DeepSeek-V3.2-Speciale, add substantial improvements in reasoning, mathematical competence and autonomous tool-use, positioning the company as a growing challenger to leading Western developers such as OpenAI and Google.

The DeepSeek-V3.2 model replaces the earlier experimental version released only weeks ago, which carried the suffix “Exp” to emphasize its developmental status. The startup now claims that the refined model matches the performance of OpenAI’s flagship GPT-5 across several high-level reasoning benchmarks. This achievement is particularly notable given DeepSeek’s open-source approach, which has increasingly demonstrated competitiveness against proprietary frontier systems from Silicon Valley.

In addition to enhanced reasoning, DeepSeek-V3.2 incorporates the ability to autonomously use external tools, including search engines, calculators and code execution modules. This integration allows the model to combine human-like problem-solving with practical digital actions. According to the company, this version is the first in its lineup to fuse internal “thinking” processes directly with tool-use capabilities, enabling it to operate in both analytical and non-analytical modes depending on the task.

DeepSeek also introduced a second model, DeepSeek-V3.2-Speciale, which is designed to excel in mathematically intensive and long-horizon reasoning challenges. The company describes this model as part of an effort to push the inference limits of open-source AI and explore the boundaries of what these systems can achieve. Early performance claims indicate that Speciale matches Google’s newly released Gemini-3 Pro and performs at gold-medal levels on elite academic contests such as the International Mathematical Olympiad and the International Olympiad in Informatics. These benchmarks have traditionally been dominated by highly specialized proprietary models.

Alongside the new models, DeepSeek announced the development of a novel training methodology for AI agents – autonomous systems capable of independently pursuing objectives, interpreting their environment, analyzing information and making decisions without continuous human oversight. The company did not provide technical details, but emphasized that this research direction is fundamental to its vision for next-generation artificial intelligence.

DeepSeek first drew global attention in January when one of its earlier models delivered unexpectedly strong performance, prompting comparisons with much larger Western competitors. The rapid release cycle of the V3.2 family suggests that Chinese AI developers are accelerating their efforts to close the technological gap and, in some domains, directly challenge U.S. leadership.

The debut of DeepSeek-V3.2 and V3.2-Speciale reinforces China’s emerging role as a serious contender in frontier AI research. As both models begin to circulate in open-source communities and enterprise testing environments, the rivalry between Chinese and U.S. developers is set to intensify, reshaping the competitive landscape of the global AI industry.

Stablecoin issuer Tether has joined a €70 million ($81.6 million) funding round for Generative Bionics, an Italy-based humanoid-robotics startup leveraging research from the Italian Institute of Technology. The investment marks another strategic step in Tether’s rapid expansion into artificial intelligence, advanced computing and deep-tech sectors.

The round was led by CDP Venture Capital, Italy’s state-backed investor, through its Artificial Intelligence Fund. Other participants include AMD Ventures, the venture arm of US semiconductor firm Advanced Micro Devices, along with several industrial backers.

According to Generative Bionics, the newly raised capital will accelerate product development, the training of physical AI systems, and the company’s move toward industrial-scale deployment. The startup will integrate 70 engineers from the Italian Institute of Technology as part of its expansion and aims to debut its first fully assembled humanoid robot at CES in Las Vegas this January.

Generative Bionics specializes in building advanced robotic systems for industrial environments, combining tactile sensing, learning architectures, and human–robot interaction technologies. Its robots are designed for applications in manufacturing, logistics and operational automation, where AI-driven adaptability is increasingly in demand.

For Tether, the investment adds to a growing list of ambitious initiatives outside its core stablecoin business. Backed by large reserves and benefiting from high interest rates, Tether expects to generate around $15 billion in profits this year, much of which is being reinvested into emerging technologies. The company has recently taken positions across commodities, AI, data infrastructure, and even brain-computer interfaces, including its backing of Blackrock Neurotech.

Tether CEO Paolo Ardoino has previously stated that the company’s expansion into artificial intelligence is aimed at building a future internet where individuals maintain stronger control over their personal data.

With its latest investment, Tether is positioning itself at the intersection of robotics and AI – sectors accelerating rapidly as automation reshapes global industry. Generative Bionics’ upcoming launch at CES will now be a closely watched milestone in that evolution.

As global demand for AI-ready networks accelerates, Nokia is reshaping its business to position itself at the center of next-generation connectivity.

Nokia Oyj is intensifying its focus on the artificial intelligence infrastructure boom, unveiling a broad restructuring plan designed to transform the Finnish telecommunications company into a key supplier of network systems for AI data centers. The shift marks one of Nokia’s most significant strategic pivots in recent years as it seeks to regain momentum in a rapidly evolving technology landscape.

Chief Executive Officer Justin Hotard said ahead of the company’s Capital Markets Day in New York that the new structure is expected to drive double-digit operating income growth over the coming years. Nokia forecast that its annual operating profit will reach between €2.7 billion and €3.2 billion by 2028, reflecting a renewed focus on efficiency and long-term profitability.

Market Reaction and AI-Driven Prospects

Despite the company’s ambitious plans, Nokia’s shares fell as much as 6.3% in Helsinki on Wednesday, marking the fifth consecutive day of declines. The selloff follows a strong surge last month, when Nvidia Corp. announced a $1 billion equity investment that gave the chipmaker a 3% stake in Nokia. Subsequent market volatility and concerns about a broader AI-driven bubble, however, weighed on the stock.

The restructuring underscores how critical AI infrastructure has become for Nokia’s future. Once the world’s most valuable mobile phone company, Nokia has struggled in recent years to prove that its mobile network division can deliver sustained growth. The explosive rise in artificial intelligence — and the massive data-center buildouts required to support it — offers Nokia a potentially transformative opportunity.

A Streamlined Structure for a New Growth Cycle

Beginning in January, Nokia will reorganize its operations into two major divisions in an effort to simplify its corporate structure and focus on areas of highest potential.

The first division, Network Infrastructure, will encompass Nokia’s optical networking systems, which provide the ultra-high-capacity fiber lines essential for moving enormous volumes of AI-related data. The second division, Mobile Networks, will retain its role as the backbone of Nokia’s radio, wireless and mobile connectivity business. This unit will also absorb the company’s patent portfolio under Nokia Technologies, as well as its cloud and network services operations.

Hotard said the shift is intended to position Nokia “directly where we see the opportunity,” adding that winning in AI requires being the leader in connectivity. The restructuring will also include the sale of four underperforming assets, among them the company’s fixed wireless access portfolio and its microwave backhaul division.

Positioning for the AI Data-Center Supercycle

Although AI and cloud clients accounted for just 6% of Nokia’s net sales in the third quarter, company executives describe the segment as the largest opportunity ahead. Demand for advanced optical networking hardware, high-performance IP routing and upgraded fiber systems is expected to soar as companies race to build the infrastructure needed for modern AI workloads.

To strengthen its capabilities, Nokia has recently acquired Infinera Corp., expanding its presence in data-center networking. It has also begun integrating Nvidia chips into next-generation 5G and 6G radio components, deepening its ties with Silicon Valley’s most valuable company. Additionally, Nokia hired Finland’s former ambassador to the U.S. to help expand its defense-related networking solutions.

The AI-focused division will continue to be led by David Heard, who is overseeing Nokia’s push into optical networking, IP switching and routing upgrades.

Mobile Networks: Strong Technology, Weak Returns

Despite its technological depth, Nokia’s mobile networks business has struggled to deliver consistent returns. Hotard acknowledged that the division has “not generated an acceptable return,” stressing that the priority is to build a more profitable business rather than simply a larger one.

After the recent departure of Tommi Uitto, Hotard will serve as interim head of the mobile networks unit. Meanwhile, Raghav Sahgal, who previously led the cloud and network services division, will assume the role of Chief Customer Officer.

A Measured Expansion into Defense

Nokia’s defense operations will continue as a standalone “incubator” unit. While the company plans to expand its presence in the defense connectivity sector, Hotard said this segment is not expected to approach the revenue scale of its core divisions in the near term.

Looking Ahead

Nokia’s restructuring marks a decisive move to align the company with the accelerating global rollout of AI infrastructure. With Nvidia’s backing, strong demand from hyperscalers and telecom operators, and a sharpened corporate focus, Nokia is positioning itself as a critical enabler of next-generation connectivity.

Whether these changes deliver the anticipated financial turnaround will depend on the company’s ability to scale its AI-centric operations, streamline execution, and capitalize on the massive investment cycle now reshaping the global technology ecosystem.

Iceye, the Finnish satellite company increasingly embedded in Europe’s defense strategy, has raised new capital at a €2.4 billion ($2.8 billion) valuation. The round, led by US venture firm General Catalyst, includes €150 million in fresh equity and €50 million in secondary sales for early investors.

CEO Rafal Modrzewski said Iceye reached profitability this year, but accelerating satellite launches has become critical amid Europe’s rapid military build-up following Russia’s invasion of Ukraine. “Nobody wants Europe to be ready in five years,” he said. “People want Europe to be ready tomorrow.”

Founded in 2014, Iceye originally focused on environmental monitoring before shifting to defense. Its synthetic aperture radar (SAR) satellites can capture high-resolution images through clouds, smoke, and darkness – capabilities increasingly sought after by governments. The company has secured deals with the US, UAE, Japan, and, more recently, a growing list of European armed forces including Poland, Portugal, Finland and the Netherlands. Two more agreements in Europe are under discussion.

To expand production capacity, Iceye formed a joint venture with German defense group Rheinmetall AG earlier this year. Rheinmetall expects up to €2 billion in satellite-related orders in upcoming quarters tied to the partnership.

General Catalyst, now one of Europe’s most active defense investors, will take a board seat through its head of Europe, Jeannette zu Fürstenberg. “It’s one of the fastest-moving segments in defense and government procurement,” she said.

Iceye launched 25 satellites this year and plans to scale to one per week next year. Revenue is expected to exceed previous forecasts after Modrzewski projected earlier that sales would double to more than €200 million in 2025.

The latest round drew institutional support from AP Møller Holding, Bpifrance, Finnish investors, and several Polish backers including BGK’s investment arm and the family office of InPost founder Rafal Brzoska. Iceye has raised €600 million to date.

With fresh capital, the company plans to strengthen its software and data-analysis tools, aiming to provide clients with real-time intelligence. “At the end of the day, nobody buys satellites just to own satellites,” Modrzewski said. “People want to understand what their enemy is doing – and what they’re about to do.”

Alphabet Inc. is closing in on Nvidia Corp. in the race for tech-market dominance as strong demand for its new Gemini AI models and growing interest in its in-house AI chips lift the company’s shares. The Google parent rose 2% on Wednesday, bringing its market value to roughly $3.9 trillion, just $300 billion shy of Nvidia’s $4.2 trillion valuation.

The rally adds to a stunning run that has seen Alphabet gain about 37% since mid-October, adding nearly $1 trillion in market value over that period. Investors are increasingly betting that Alphabet’s advances in both AI software and hardware could reshape the competitive landscape.

A major catalyst came from a report that Meta Platforms Inc. is in discussions to start using Google’s tensor processing units – or TPUs – in its data centers beginning in 2027. Meta may also rent TPU capacity through Google Cloud next year. Such a partnership would give Alphabet’s chips a high-profile customer and challenge Nvidia’s position as the dominant supplier of AI accelerators.

The prospect of real competition sent Nvidia shares down 5.1%, with rival chipmaker AMD falling 7.8%. Nvidia’s valuation has cooled to 26 times forward earnings, well below its long-term average. Alphabet, boosted by the excitement surrounding Gemini, now trades at 27 times forward earnings, up sharply from past levels.

Analysts say the developments highlight a shift in market sentiment. Alexandra Morris, investment director at Skagen AS, said the belief that Nvidia alone would supply chips for the global buildout of AI data centers “has changed,” though she added that TPUs still need to prove they can match Nvidia’s efficiency and compute performance at scale.

The surge in Alphabet shares has pushed its relative strength index to about 75, a level many technical analysts consider overbought. Matthew Maley of Miller Tabak noted that Alphabet could assume a “leadership role” in the AI industry but may also be due for a near-term pullback.

Even so, the momentum reflects growing expectations that Alphabet may play a much larger role in the AI hardware market than previously assumed – potentially setting up a reshaping of the global tech leadership rankings.

Uber Technologies Inc. has begun offering fully driverless rides in partnership with WeRide Inc. on Yas Island in Abu Dhabi, marking a significant expansion of autonomous mobility in the Middle East. The companies launched their initial pilot with safety operators in late 2024. Now, a year later, select UberX and Uber Comfort riders can be matched with a completely driverless vehicle across a 12–square-mile area.

Riders can also choose a new “Autonomous” option in the Uber app, which increases the likelihood of receiving a robotaxi and is priced similarly to Uber Comfort. Trips beyond Yas Island, including routes from Abu Dhabi International Airport and other surrounding islands, will continue to operate with a safety operator for the time being. This broader semi-autonomous service covers approximately 30 square miles.

Uber and WeRide plan further expansion within Abu Dhabi and intend to bring driverless operations to Dubai in the near future. Their partnership extends well beyond the UAE. Earlier this year, the companies agreed to launch in 15 additional cities worldwide, including Riyadh, where service with safety operators has already started. In the United States, Uber currently offers driverless rides through an existing partnership with Alphabet Inc.’s Waymo in Phoenix, Austin, and Atlanta.

Since selling its in-house autonomous division in 2020, Uber has shifted toward establishing itself as a global platform for autonomous mobility, investing heavily in technology partners such as WeRide and forging more than 20 partnerships focused on AV deployment. The company said that dozens of Uber-managed WeRide vehicles are currently operating in Abu Dhabi and that the fleet will expand with rising demand.

WeRide, whose shares trade in both the US and Hong Kong, recently reported a third-quarter net loss of 307 million yuan ($43 million), an improvement from a 1.04-billion-yuan loss a year earlier. The company now holds autonomous driving permits in eight countries as it accelerates international growth.

For the UAE, the launch of fully driverless rides on Yas Island reinforces the country’s ambition to become a leading global hub for smart mobility. With further expansion planned across Abu Dhabi and Dubai, the region is positioning itself at the forefront of the autonomous transportation industry.