Technology News Today – The Latest in Tech, AI & Startup News, December 9, 2025

1. Microsoft commits $5.4B to AI infrastructure expansion in Canada

Microsoft is ramping up its global AI infrastructure build-out, pledging to invest more than C$7.5 billion (about $5.4 billion) in Canada over the next two years. The tech giant plans to add several new data centers in Quebec and other provinces, significantly expanding its cloud and AI capacity in the country. The investment includes new AI-focused data centers, training programs to upskill workers on AI and cloud technologies, and collaborations with local institutions to support AI research and digital innovation.

This is part of a broader trend in which hyperscale cloud and AI leaders are racing to build out compute capacity near major enterprise and government customers, both to meet surging demand for AI workloads and to address data residency and sovereignty concerns. For Canada, the move reinforces the country’s positioning as a regional AI and tech hub and signals that competition between Microsoft, Amazon, and Google Cloud for AI workloads is intensifying at the infrastructure layer.

Why It Matters: This is another clear signal that AI infrastructure is becoming the new strategic battleground for Big Tech — and that countries with stable regulation and talent pools can attract multi-billion-dollar AI and cloud investments.

Source: Reuters.

2. Brookfield and Qatar launch $20B AI infrastructure joint venture

Brookfield Asset Management and Qatar’s state-backed QAI have formed a massive $20 billion joint venture focused on AI infrastructure, targeting data centers and related energy-intensive assets worldwide. The venture will fund, build, and operate facilities that power large-scale AI and cloud workloads, including high-performance computing sites and hyperscale data centers. The partners are positioning the platform as a long-term owner of “AI infrastructure as an asset class,” tying together real estate, energy, and digital compute.

The deal underscores how infrastructure investors and sovereign capital are pouring money into the physical backbone of the AI boom — power-hungry data centers, cooling, and grid connections. Rather than betting on any single AI model or startup, Brookfield and Qatar are effectively selling “picks and shovels” to the AI gold rush, capturing returns from long-term contracts with cloud providers and Big Tech. It also highlights that capital availability for AI-related infrastructure remains deep, even as some late-stage AI startup valuations come under greater scrutiny.

Why It Matters: AI infrastructure is becoming its own asset class, attracting sovereign and infrastructure megafunds and reinforcing that the real bottlenecks in AI are power, land, and data center capacity —not just model quality.

Source: Reuters.

3. EU opens antitrust probe into Google’s use of online content for AI training

European Union regulators have launched a fresh antitrust investigation into Alphabet’s Google over how it uses publishers’ articles and YouTube videos to train its AI models and power products such as AI Overviews. Authorities are probing whether Google’s practices unfairly exploit news publishers’ and creators’ content without adequate consent or compensation, potentially breaching EU competition rules. The probe also examines how Google integrates AI features into search and whether this behavior disadvantages rival services or content providers.

This investigation lands as publishers and media groups worldwide push back against AI firms that scrape or license content to train large models. In the EU, where regulators have already taken aggressive action on privacy and platform dominance, this case could set a precedent for how tech and AI companies must negotiate with rights holders when building foundation models. Depending on the findings, Google could face fines, behavioral remedies, or new obligations to negotiate licensing terms with publishers and creators across the bloc.

Why It Matters: The case could reshape how AI and search platforms use news and creator content in Europe — and may set a de facto global standard for AI training, licensing, and compensation rules.

Source: Reuters / Bloomberg.

4. US clears Nvidia’s H200 AI chips for export to China, shifting chip-control strategy

The United States has decided to allow Nvidia to export its H200 AI accelerators — its second-fastest data-center chips — to China, marking a notable shift in Washington’s tech-controls strategy. According to reports and official comments, the approval follows negotiations in which the Trump administration agreed to permit shipments in exchange for a 25% tariff on the advanced AI chips. The move partially relaxes earlier export curbs that had blocked Nvidia’s highest-end AI hardware from reaching Chinese hyperscalers and tech companies.

The decision highlights the balancing act between national security and economic interests: US policymakers want to constrain China’s access to the very top of AI compute while still allowing some trade in slightly downgraded chips under controlled conditions. For Nvidia, the H200 ruling reopens a major growth market at a time when Chinese demand for AI hardware remains intense, even as domestic Chinese chipmakers race to catch up. For Beijing and Chinese AI startups, the move relieves some short- to medium-term pressure but also reinforces that access to top-tier Western AI chips will remain contingent on geopolitical negotiations.

Why It Matters: Easing restrictions on Nvidia’s AI chips signals that AI export controls are moving from blanket bans to more nuanced, tariff-driven regimes — reshaping the competitive landscape for US and Chinese AI players.

Source: Bloomberg / Reuters

5. China’s homegrown AI chips go head-to-head with Nvidia’s H200

A separate Reuters analysis digs into how new Chinese AI chips from firms like Huawei and Biren compare with Nvidia’s H200 accelerator, which has become a benchmark for high-end AI compute. While Nvidia still leads in overall performance and ecosystem maturity, Chinese vendors are closing the gap on metrics such as FLOPS, memory bandwidth, and power efficiency, particularly for training and inference of large AI models tailored to Chinese-language and regional workloads. Some Chinese accelerators are designed to circumvent US export controls by operating just below certain performance thresholds.

The piece notes that Chinese cloud providers and AI startups are increasingly testing and deploying domestic silicon to reduce reliance on US hardware and hedge against future sanctions. At the same time, Nvidia’s software stack (CUDA, libraries, and tooling) remains a major moat, meaning Chinese AI developers must invest in alternative frameworks and optimization layers. In the long term, this competition could split the AI infrastructure market into partially incompatible ecosystems—one centered on US chips and tooling, the other on China’s domestic stack.

Why It Matters: The race between Chinese AI chips and Nvidia’s flagship accelerators will shape global AI capacity, pricing, and supply chains — and determines how much leverage US export controls retain over China’s AI ambitions.

Source: Reuters.

6. Apple stock pops as investors rotate out of the AI trade

Apple shares are rallying after a tough year in which investors repeatedly criticized the company for lacking a bold AI strategy. A new Bloomberg analysis argues that as Wall Street grows more cautious about an “AI bubble,” Apple’s slower, more measured AI messaging looks less like a weakness and more like a defensive asset. While Nvidia, Microsoft, and other AI-centric tech names have been under heavier scrutiny, Apple’s valuation has benefited from investors rotating into perceived “safer” Big Tech names with diversified revenue and less AI hype priced in.

The piece highlights how Apple has been quietly integrating AI into its core products — from on-device intelligence in iPhones to productivity and photo features — without relying too heavily on AI marketing buzz. That posture may now be paying off as traders reevaluate overvalued AI leaders and seek exposure to high-quality tech names that can benefit from AI without being entirely defined by it. It also underscores a broader shift in which investors are distinguishing among AI infrastructure plays, application startups, and diversified platforms with multiple profit engines.

Why It Matters: Apple’s rebound shows that “AI optionality” plus balance-sheet strength may be more attractive to investors than pure AI speculation. This signal could reshape capital flows across the tech sector in 2026.

Source: Bloomberg / Inc.

7. NextEra, America’s biggest power utility, pivots its grid strategy for AI data centers

America’s largest power company, NextEra Energy, is retooling its strategy to serve the explosive electricity demand from AI and cloud data centers. A Bloomberg opinion piece details how the utility is prioritizing grid upgrades, a renewables build-out, and transmission projects tailored to hyperscale data center customers. With AI workloads driving unprecedented power demand, NextEra is positioning itself as a key partner for Big Tech, offering long-term clean-energy contracts and grid services to support dense clusters of GPU-hungry AI facilities.

The shift illustrates how AI is now a structural force in the energy sector, influencing where new power plants get built, what mix of renewables and gas is financed, and how regulators think about reliability. Data centers for AI and cloud are increasingly competing with industrial users and cities for limited grid capacity, forcing utilities to prioritize which projects they connect and when. For investors, NextEra’s pivot also highlights a new thesis: owning the “AI grid” could become as important as owning the AI chips that sit on top of it.

Why It Matters: As AI reshapes electricity demand curves, utilities like NextEra are becoming quiet kingmakers — determining which AI and tech projects get the power they need and at what cost.

Source: Bloomberg.

8. Cybersecurity startup Saviynt raises $700M to secure AI-era identity and access

Identity and access management startup Saviynt has raised a massive $700 million in an outsized Series B round, according to a Wall Street Journal report. The funding underscores how demand is surging for tools that secure which humans — and software robots — can access critical systems as AI and automation proliferate across enterprises. Saviynt’s platform helps large organizations manage identities, enforce granular access policies, and detect anomalous behavior across cloud, SaaS, and on-prem environments, including AI-powered services.

The article notes that AI adoption is forcing companies to rethink identity and access at scale, as machine accounts, bots, and AI agents increasingly interact with sensitive data and production systems. This creates a larger attack surface and increases the risk of misconfigurations. Investors are betting that Saviynt can be a core part of the AI security stack, sitting alongside endpoint protection, data security, and cloud workload defenses. The size of the round and the late-stage valuation also indicate that cybersecurity, particularly identity-centric security, remains among the most resilient segments of the tech funding market.

Why It Matters: As AI agents and automation spread across enterprises, identity and access management becomes a critical control plane—and investors clearly see IAM startups like Saviynt as foundational security infrastructure.

Source: Wall Street Journal.

9. Anthropic and Accenture sign three-year AI services deal for enterprise clients

Anthropic, the AI startup behind the Claude models, has inked a three-year partnership with consulting giant Accenture to bring AI services to large enterprises. Under the deal, Accenture will integrate Anthropic’s AI models into customized solutions for business clients, providing AI-powered automation, knowledge management, and decision-support tools. The Wall Street Journal reports that the partnership aims to help companies move beyond pilots and proofs of concept into scaled deployments, leveraging Accenture’s consulting and integration capabilities with Anthropic’s AI technology.

The collaboration is emblematic of a growing trend: AI foundation-model startups partnering with systems integrators and consulting firms to reach conservative, highly regulated industries that are reluctant to engage directly with startups. For Anthropic, it’s a distribution and monetization play that complements its direct API business. At the same time, for Accenture, it’s another way to fill its AI portfolio alongside rival offerings from Microsoft, Google, and others. The deal also highlights how “AI services” is emerging as a huge revenue line for consultancies as enterprises seek partners that can handle governance, change management, and integration — not just APIs.

Why It Matters: This deal shows that the next phase of AI adoption in the enterprise will run through services giants like Accenture — and that AI startups will increasingly rely on these partnerships to turn model strength into real revenue.

Source: Wall Street Journal.

10. SoftBank and Nvidia eye $1B+ investment in robotics AI startup Skild AI at $14B valuation

SoftBank Group and Nvidia are in talks to lead a more than $1 billion investment in Skild AI, a robotics-focused AI startup that builds foundation models for robots rather than hardware, according to TechCrunch and Reuters. The prospective deal would value Skild AI at around $14 billion, nearly tripling its valuation from a Series B earlier this year. Founded in 2023 by former Meta AI researchers and already backed by Amazon and Lightspeed, Skild AI trains large models that aim to give robots human-like perception and decision-making across a variety of tasks and environments.

The potential investment underscores how hot the “general-purpose robotics” thesis has become—even though experts say truly versatile humanoid and mobile robots could still be years away from mainstream deployment. For Nvidia, the deal would deepen its presence in the robotics AI stack, creating a flagship customer for its AI chips and robotics SDKs. For SoftBank, it fits a long-running bet on robotics spanning past investments in companies like Boston Dynamics and various warehouse automation players. A successful close before year-end would also mark one of the largest AI startup funding rounds of 2025.

Why It Matters: If finalized, this round would cement Skild AI as a flagship player in general-purpose robotics and reinforce Nvidia’s strategy of backing AI startups that generate massive demand for its compute.

Source: TechCrunch / Reuters.

11. Unconventional AI raises $475M seed round at $4.5B valuation to redesign computing

In one of the wildest seed rounds of the year, Unconventional AI has raised $475 million at a reported $4.5 billion valuation — just two months after launch. The startup, founded by former Databricks AI leaders, is building a radically different full-stack computing architecture optimized for AI workloads, according to Tech Funding News. Its approach focuses on tightly integrating optical interconnects and specialized hardware to significantly increase bandwidth and reduce latency between AI accelerators, helping overcome bottlenecks in current GPU-centric setups.

Investors are betting that the next leap in AI performance won’t just come from bigger models or more GPUs, but from rethinking how data moves through AI infrastructure. With AI training runs costing hundreds of millions of dollars, any improvement in efficiency or throughput has enormous economic leverage. The scale of the seed round — and the sky-high valuation — also reflect how much capital is chasing “picks and shovels” at the deepest layer of AI infrastructure, from networking to memory to data movement. For startup founders, it’s another sign that “AI infra” remains one of the few areas where mega-rounds are still possible.

Why It Matters: This funding underscores that investors see AI infrastructure innovation — not just new models or apps — as a multi-billion-dollar frontier, and that breakthroughs in compute architecture could reset the performance bar for the entire AI ecosystem.

Source: Bloomberg

12. Space-tech startup AnySignal raises $24M Series A for next-gen radio systems

California startup AnySignal has closed a $24 million Series A round to scale production of its advanced radio systems for satellite and defense applications, according to space industry outlet Payload. The company builds high-performance RF hardware and software that enable more flexible, software-defined communications in space, allowing satellites and spacecraft to adjust frequencies, protocols, and mission profiles dynamically. With this round, led by Upfront Ventures with participation from BlueYard Capital and others, AnySignal plans to move into a larger manufacturing facility and expand its engineering team.

The funding reflects how “new space” is increasingly about software, signal processing, and flexible architectures rather than launch vehicles. As space becomes more crowded with commercial constellations and defense assets, reliable, adaptable communications are becoming a critical part of the infrastructure stack—akin to networking in terrestrial cloud data centers. AnySignal’s tech sits at the intersection of space, defense, and deep tech, a zone where governments and large contractors are eager to work with nimble startups that can iterate faster than traditional primes.

Why It Matters: Space is becoming a software-defined network — and startups like AnySignal, which modernize radio and RF technologies, are poised to be core suppliers in both commercial constellations and defense space programs.

Source: Payload.

13. Petco data breach exposes Social Security numbers and IDs in major leak

Petco, a pet retail giant, has disclosed a major data breach that exposed sensitive customer information, including Social Security numbers, driver’s license details, and financial data. According to The Tech Buzz, the incident stemmed from a significant security lapse that left highly sensitive records accessible and subsequently exploited, resulting in a large-scale leak. The company is now notifying affected customers, working with cybersecurity experts, and may face heightened regulatory scrutiny given the nature of the data exposed.

The breach underscores that even as companies invest in AI, cloud, and digital experiences, basic security hygiene and access controls can still fail in surprisingly simple ways. Regulated identifiers like SSNs and government IDs are particularly sensitive, triggering obligations under state breach notification laws and potentially inviting class-action lawsuits. For the broader tech and retail sector, the case is a reminder that data minimization, encryption, and strict access governance are not optional — especially as more customer data gets fed into analytics and AI systems that expand the attack surface.

Why It Matters: This is another high-profile example that poor security around legacy data can undermine even sophisticated tech initiatives — and regulators are increasingly unwilling to treat such breaches as mere “IT issues.”

Source: Yahoo News via TechCrunch

14. Pharma firm Inotiv confirms Qilin ransomware breach affecting thousands

Pharmaceutical research company Inotiv has confirmed a significant ransomware attack and data breach carried out by the Qilin group, according to a recent SEC filing and independent cyber reports. The attack, detected in August 2025, disrupted access to key systems and has now been linked to the exposure of personal, financial, and health information of more than 9,000 individuals. In a December disclosure, Inotiv said it has restored systems, completed its internal investigation, and is coordinating with law enforcement while notifying affected patients, employees, and partners.

Security analysts note that the incident highlights growing supply-chain and research-related cyber risk in the life sciences sector, where companies hold valuable clinical, preclinical, and patient data. As pharma and biotech firms increase their use of AI for drug discovery, lab automation, and clinical analytics, their data estates are becoming more attractive targets for ransomware operators. The Inotiv breach fits a pattern in which attackers aim at mid-sized research organizations that may not have the same security budgets as Big Pharma but still hold highly monetizable data.

Why It Matters: The Inotiv ransomware attack is a warning shot for pharma and biotech — AI-driven research is only as resilient as the cybersecurity protecting the data behind it.

Source: HIPAA Journal / Cyber News Centre.

15. Australia to enforce under-16 social media ban, signaling a global shift in online youth policy

Australia is set to become the first country to implement a nationwide minimum age for social media, with a law restricting most platforms to users 16 and older taking effect this week. Reuters reports that the law requires major tech and social media companies to verify ages and block users under 16, with stiff penalties for non-compliance. Officials say the move is aimed at reducing mental-health harms, cyberbullying, and exposure to inappropriate content among children, while critics warn about privacy, enforcement challenges, and the risk of pushing teens to unregulated platforms. Reuters

This policy adds to mounting pressure on Big Tech over how it handles youth safety and data globally. Regulators in the EU, US, and elsewhere are closely monitoring whether Australia’s approach is workable in practice, particularly given the technical and civil-liberties debates surrounding age-verification systems. AI and age-estimation tech will likely play a central role in how platforms attempt to comply, raising new questions about biometric data, accuracy, and bias. If Australia’s model proves politically popular, it may accelerate similar legislation in other democracies.

Why It Matters: Australia’s under-16 social media ban could become a blueprint — or a cautionary tale — for how governments regulate youth access to tech platforms and deploy AI for age verification.

Source: Reuters.