DD
MM
YYYY

PAGES

DD
MM
YYYY

spot_img

PAGES

Home Blog Page 4

Amazon Cuts Jobs In AGI Unit As It Sharpens AI Strategy While Ramping Up $200bn Investment

0

Amazon has laid off employees in its artificial general intelligence (AGI) organization as the technology giant refines its artificial intelligence strategy.

The move underscores how even the industry’s biggest AI investors are reallocating talent while committing unprecedented sums to AI infrastructure.

The company confirmed the job cuts on Wednesday but did not disclose how many employees were affected or identify the specific teams impacted within the AGI division. The unit is responsible for developing Amazon’s frontier AI models and also houses teams working on custom AI silicon and quantum computing, two technologies viewed as critical to the company’s long-term AI ambitions.

The restructuring comes as Amazon balances aggressive investment in AI with continued efforts to streamline operations following years of workforce reductions.

“This is a fast-moving space, and we’re sharpening our focus on the initiatives that matter most for customers, so we can move faster on what counts,” an Amazon spokesperson said in a statement.

“That focus means some difficult decisions, including eliminating some roles within parts of our AGI organization, even as we continue to invest in the areas most important to our customers’ future.”

Reuters first reported the layoffs.

The latest reductions suggest that AI spending does not necessarily translate into broad-based hiring. Instead, major technology companies are increasingly reallocating resources toward projects with the greatest commercial potential, trimming overlapping teams while expanding investment in core infrastructure, advanced models and specialized engineering talent.

Amazon has been engaged in a multiyear cost-cutting effort since the post-pandemic slowdown prompted large technology companies to reassess their workforce needs. Since late 2022, the company has eliminated more than 30,000 jobs across multiple divisions, including devices, cloud computing, advertising, communications and entertainment. Smaller rounds of layoffs have continued throughout 2026 as Amazon seeks to improve operating efficiency while redirecting capital toward AI.

The AGI division sits at the center of Amazon’s strategy to compete with industry leaders such as OpenAI, Anthropic and Google in the race to build increasingly capable foundation models. Artificial general intelligence generally refers to AI systems capable of matching or surpassing human performance across a broad range of cognitive tasks, though no company has yet achieved that milestone.

Amazon entered the frontier model race later than some rivals but has accelerated development over the past two years. In 2024, the AGI organization introduced its Nova family of foundation models, designed to support enterprise customers through Amazon Web Services and power generative AI applications across Amazon’s businesses.

The unit underwent a major leadership overhaul last December when Amazon appointed longtime AWS executive Peter DeSantis to lead the organization, replacing Rohit Prasad. The leadership change signaled Amazon’s intention to integrate AI model development more closely with its cloud infrastructure strategy and accelerate commercialization of its AI technologies.

The organization has also experienced executive turnover. In February, David Luan, who headed Amazon’s AGI lab after joining through the acquisition of startup Adept in 2024, left the company, raising questions about leadership continuity as Amazon pushes to narrow the gap with more established AI competitors.

Despite the layoffs, Amazon said AI remains one of its highest strategic priorities. The company has been building large AI models for several years, and “it remains one of the most important things we’re working on,” the spokesperson said.

DeSantis acknowledged in an interview with CNBC last month that Amazon still trails the industry’s most advanced AI developers in certain frontier capabilities.

“Our models haven’t been at the very frontier for the very largest, most demanding workloads,” he said, adding that Amazon is working to strengthen its model portfolio with the goal of developing one of the “most capable intelligent models out there.”

He thus confirmed Amazon’s recognition that while it possesses one of the world’s largest cloud computing platforms and extensive AI infrastructure, it has yet to establish the same reputation for cutting-edge foundation models enjoyed by competitors including OpenAI, Anthropic and Google.

Amazon’s plan centers on leveraging its unique competitive advantages rather than competing solely on model performance. Through AWS, the company offers customers access to multiple third-party models, including Anthropic’s Claude family, alongside its own Nova models, allowing enterprises to choose among different AI systems while keeping workloads within Amazon’s cloud ecosystem.

The workforce reductions come as Amazon prepares to report second-quarter earnings next week, when investors are expected to closely scrutinize AI-related spending, cloud growth and returns on the company’s massive capital investments.

Amazon has projected capital expenditures of approximately $200 billion this year, representing an increase of more than 50% from 2025. The spending will primarily fund AI data centers, custom Trainium and Inferentia chips, networking infrastructure and expanded cloud capacity required to train and deploy increasingly sophisticated AI models.

To support those investments, Amazon has also raised tens of billions of dollars through debt markets, joining Microsoft, Alphabet and Meta in making record capital commitments to artificial intelligence.

EU Clears $110bn Paramount-Skydance Merger With Warner Bros. Discovery, But U.S. Court Battle Clouds Closing Timeline

0

European Union antitrust regulators on Wednesday approved Paramount Skydance’s proposed $110 billion acquisition of Warner Bros. Discovery, removing one of the final major international regulatory hurdles for what would become one of the largest media mergers in recent years.

The clearance comes with legally binding concessions from Paramount aimed at preserving competition in Europe’s film distribution market. However, while the decision represents a significant boost for the companies, the merger now faces its biggest test in the United States, where a lawsuit from a coalition of state attorneys general has temporarily halted progress toward closing the transaction.

Investors welcomed the European approval, sending Paramount shares about 3% higher in midday trading.

The European Commission said it approved the transaction after Paramount agreed to divest its stake in United International Pictures (UIP), a long-standing European film distribution joint venture. The company also committed not to enter into any film distribution agreement with Universal Pictures in Europe for the next decade.

According to the Commission, the commitments fully resolve concerns that the merged company could coordinate distribution activities with Universal or Disney, potentially reducing competition for theatrical releases across Europe.

“These commitments fully address the competition concerns identified by the Commission by ensuring that the films of the merged entity will not be distributed jointly with those of Universal or Disney,” the Commission said.

The remedies highlight regulators’ increasing scrutiny of distribution networks, particularly as Hollywood studios seek greater scale to navigate slowing box office recovery, rising production costs and intensifying competition from global streaming platforms.

If completed, the merger would reshape the global entertainment industry by combining two of Hollywood’s most valuable film studios and some of the world’s best-known television and streaming brands. The combined company would control Paramount Pictures and Warner Bros. Pictures, alongside streaming platforms Paramount+ and HBO Max, while bringing together premium television assets including CBS, CNN, Discovery Channel, Nickelodeon, TNT Sports, Cartoon Network, HGTV, Food Network and a vast library of television and film content.

The transaction is designed to create a media giant with greater financial scale, stronger bargaining power with advertisers and distributors, and a broader direct-to-consumer streaming business capable of competing more effectively against technology-backed rivals such as Netflix, Amazon and Apple, all of which have dramatically increased spending on original entertainment.

Industry analysts have long argued that consolidation has become necessary as traditional television revenues decline and studios struggle to generate consistent profits from streaming, forcing media companies to seek larger subscriber bases and greater operating efficiencies.

The merger has already secured approval from the U.S. Department of Justice’s Antitrust Division, as well as regulators in several other jurisdictions, suggesting that competition authorities outside California have concluded the transaction can proceed with appropriate safeguards.

The principal obstacle now lies in the U.S. courts.

Last week, a coalition of state attorneys general led by California Attorney General Rob Bonta filed a lawsuit seeking to block the merger on antitrust grounds. The states argue that combining two major Hollywood studios, extensive television assets, and leading streaming platforms would significantly increase market concentration and reduce competition across film production, television programming, content licensing, and digital streaming.

Earlier this week, a California district judge issued a temporary restraining order preventing the companies from taking additional steps toward completing the merger for 14 days while the court considers whether to grant a longer injunction.

Although the order is procedural and does not determine the merits of the case, it injects fresh uncertainty into the timetable for closing the transaction. Before the lawsuit, Paramount had maintained that it remained on track to complete the merger by the end of September.

Some analysts believe that achieving that target will depend largely on how quickly the California litigation is resolved. If the court extends the injunction or allows the case to proceed to a full trial, the companies could face months of additional legal uncertainty despite having cleared nearly every major regulatory review worldwide.

U.S. Senate Panel Advances Bill Targeting Chinese Automakers, Raises Questions Over Mercedes-Benz Ownership

0

A bipartisan U.S. Senate committee has advanced legislation designed to tighten restrictions on Chinese automakers operating in the United States, but lawmakers acknowledged the proposal could unintentionally ensnare one of Germany’s largest luxury carmakers, Mercedes-Benz, because of its Chinese shareholders.

The Senate Commerce Committee on Wednesday approved the Motor Vehicle Modernization Act of 2026, legislation aimed at strengthening barriers against Chinese-linked vehicle manufacturers and connected vehicle technologies over national security concerns.

However, during the committee’s markup, Chairman Ted Cruz, a Republican from Texas, warned that the bill’s current language could inadvertently prevent Mercedes-Benz from selling vehicles in the U.S., highlighting the growing complexity of efforts to separate Western industries from Chinese investment.

“We would never consider” banning Mercedes-Benz, Cruz said, noting that the legislation would need to be revised before becoming law.

The concern centers on a provision that would prohibit companies with at least 15% Chinese ownership from selling vehicles in the United States. Mercedes-Benz exceeds that threshold because its two largest individual shareholders are Chinese investors.

Chinese state-owned automaker BAIC (formerly Beijing Automotive Industry Corporation) owns a 9.98% stake in Mercedes-Benz Group, while Geely founder Li Shufu holds another 9.69%. Together, the two investments amount to nearly 20% of the German automaker’s shares.

Over the past two decades, Chinese companies and investors have accumulated minority stakes in numerous Western manufacturers, creating challenges for governments seeking to restrict Chinese influence without disrupting longstanding commercial relationships.

The proposed legislation seeks to codify and expand federal restrictions designed to keep Chinese-linked vehicle technology out of the U.S. market. Policymakers from both parties have argued that modern connected vehicles collect vast amounts of data through cameras, microphones, GPS systems and wireless communication technologies, creating potential national security risks if the data can be accessed by foreign adversaries.

Washington has steadily expanded scrutiny of Chinese involvement in strategic industries, including semiconductors, telecommunications, artificial intelligence and electric vehicles, amid broader geopolitical competition between the United States and China.

Supporters of the legislation say the measure is intended to protect both national security and America’s manufacturing base.

“We’re preventing an absolute, total, and complete destruction of our industrial base,” said Senator Bernie Moreno, an Ohio Republican who introduced the bill alongside Democratic Senator Elissa Slotkin of Michigan.

The bipartisan sponsorship underscores growing consensus in Congress that Chinese participation in sensitive industries warrants stricter oversight, even as lawmakers continue to debate how broadly those restrictions should apply.

Mercedes-Benz has previously declined to comment directly on the proposed legislation but emphasized its significant footprint in the United States. The company employs more than 10,000 people across the country and operates major manufacturing facilities in Alabama and South Carolina, producing vehicles both for domestic consumers and export markets.

Those investments have made Mercedes-Benz one of the largest foreign automotive manufacturers operating in the United States, raising questions about how ownership-based restrictions should be applied to companies with substantial American operations.

Seeking to ease concerns, Moreno told lawmakers that Mercedes-Benz would have until 2030 to comply with the proposed ownership threshold and could also apply for a waiver if necessary.

The provision suggests Congress may seek to provide flexibility for companies with significant U.S. economic contributions while maintaining pressure to reduce Chinese ownership in strategically important sectors.

The debate also exposed competitive tensions within the U.S. auto industry.

During the committee session, Cruz accused General Motors of supporting the ownership provision because it would disadvantage Mercedes-Benz and strengthen the competitive position of GM’s luxury Cadillac brand.

“GM is pushing for this provision to get Mercedes-Benz out of the market,” Cruz said.

General Motors remains the best-selling automaker in the U.S. market.

The proposed legislation comes as the U.S. government continues to tighten restrictions on Chinese automotive technology. Federal officials have focused more on connected vehicles, warning that software-enabled cars capable of collecting location data, biometric information and communications could present intelligence and cybersecurity risks if developed or controlled by companies with links to foreign governments.

In the U.S. industrial policy, national security considerations are playing a growing role in trade, investment and manufacturing decisions. While the primary target this time is Chinese automakers and technology suppliers, the Mercedes-Benz debate shows how deeply integrated global ownership structures have become.

However, lawmakers are expected to revise the bill’s ownership provisions to ensure that restrictions aimed at limiting Chinese influence do not unintentionally affect long-established international manufacturers with significant operations, investments and employment in the United States.

Trump Expands AI Power Cost Pledge to Shield Households as Data Center Boom Strains U.S. Electric Grid

0

President Donald Trump is expected to unveil an expanded electricity affordability initiative on Thursday aimed at preventing American households from bearing the cost of the country’s rapidly growing artificial intelligence industry.

The move comes as his administration accelerates efforts to cement U.S. leadership in AI while addressing mounting concerns over the sector’s impact on the nation’s power grid.

The updated Ratepayer Protection Pledge broadens an initiative first introduced in March, committing utilities, governors, lawmakers and data center developers to ensure that companies building and operating AI data centers pay electricity rates above standard tariffs so that infrastructure and energy costs are not shifted onto residential consumers.

According to a White House official, Trump will announce the expanded pledge alongside Energy Secretary Chris Wright, Environmental Protection Agency Administrator Lee Zeldin and several governors who endorsed the original agreement. The White House said the voluntary, nonbinding pledge will now apply to electricity systems serving approximately 80% of the power delivered to U.S. homes and businesses, substantially expanding its geographic reach.

The initiative comes as the explosive growth of artificial intelligence places unprecedented pressure on America’s electricity infrastructure. Technology companies including Microsoft, Amazon, Google, Meta and OpenAI-backed infrastructure projects are investing hundreds of billions of dollars in new AI data centers, facilities that consume enormous amounts of electricity to power advanced semiconductors and cooling systems.

Industry analysts estimate that AI-related electricity demand could become one of the fastest-growing sources of power consumption in the United States over the next decade, forcing utilities to accelerate investments in generation, transmission and grid modernization. The rapid expansion has raised concerns among consumer advocates, regulators and policymakers that households could ultimately shoulder part of the financial burden if utilities recover the costs of new infrastructure through higher electricity rates.

Trump’s expanded pledge seeks to address those concerns by reinforcing the principle that hyperscale data center operators should bear the incremental costs associated with their exceptionally large electricity demand rather than spreading those expenses across ordinary residential customers.

The challenge is acute because America’s electricity network was not designed to accommodate the unprecedented concentration of demand created by modern AI facilities.

Large AI data centers can consume as much electricity as medium-sized cities, requiring new substations, transmission lines and generating capacity. In many regions, utilities are struggling to connect these facilities quickly as electricity demand begins to grow after years of relatively flat consumption.

Those pressures are becoming increasingly visible across regional electricity markets.

The PJM Interconnection, the nation’s largest regional transmission organization, which supplies electricity to about 67 million people across 13 states and the District of Columbia, has become one of the clearest examples of the mounting strain on the grid. Transmission congestion costs in PJM surged 81% to $3.2 billion in 2025, highlighting growing bottlenecks as existing infrastructure struggles to accommodate rising electricity demand.

The challenges intensified last week when PJM’s latest annual capacity auction produced record-high clearing prices, reflecting tightening electricity supplies and growing concerns about whether sufficient generation capacity will be available to meet future demand.

The Federal Energy Regulatory Commission (FERC) is scheduled to meet Thursday to discuss PJM’s increasingly fragile supply-demand balance, an issue that has become central to the national debate over AI-driven electricity consumption.

Energy policy analysts argue that the existing structure of the U.S. electricity market may be ill-equipped to accommodate the needs of both traditional consumers and hyperscale computing facilities.

“Today’s electricity system forces almost everyone, from homeowners to hyperscale data centers, to depend on the same network,” Travis Fisher, director of energy and environmental policy studies at the Cato Institute, recently wrote.

“That one-size-fits-all model increasingly does not serve either group well. Large customers may wait a decade or longer for service, while ordinary ratepayers worry that expanding the grid for massive new industrial loads will ultimately increase their own bills.”

Those concerns have fueled calls for new regulatory approaches, including dedicated transmission infrastructure, direct power purchase agreements and behind-the-meter generation that would allow large data centers to secure electricity supplies without placing additional pressure on shared distribution systems.

The data center industry, however, argues that it is being unfairly blamed for electricity price increases that stem from broader structural challenges.

Industry groups contend that AI investment is catalyzing long-overdue upgrades to America’s aging electricity infrastructure while pointing to several other factors driving higher power costs, including the retirement of conventional power plants, delays in building new transmission lines, rising construction costs and increasing demand from electrification across multiple sectors of the economy.

Supporters also believe that the billions of dollars flowing into data center construction are stimulating local economies through investment, employment and tax revenues while strengthening America’s competitive position in artificial intelligence.

The enormous energy requirements of AI are creating new challenges for electricity markets, utilities and regulators. The Trump administration now has to ensure that affordable electricity remains available for households while enabling rapid expansion of AI infrastructure, making it a key component of its industrial and technology strategy.

Although the expanded Ratepayer Protection Pledge is voluntary and does not create legally enforceable obligations, the White House hopes it will establish an industry standard that encourages utilities and data center developers to structure electricity contracts in ways that insulate residential customers from AI-related infrastructure costs.

Why Africa’s Next Tech Growth May Come From Solving Everyday Problems

0

Big tech stories often focus on shiny ideas. Artificial intelligence. Space projects. Large data centers. New apps with clever names. These things matter, but Africa’s next wave of tech growth may come from something much closer to daily life. It may come from fixing the problems people face every morning.

Everyday Problems Create Real Markets

A good tech idea does not need to sound futuristic. It needs to solve something people care about. In Africa, many of those needs are clear because they show up in normal routines.

A trader needs fast payment. A farmer needs weather updates and market access. A patient needs medicine that is real and available. A driver needs a better way to move goods through traffic. These are practical problems, and practical problems can create strong businesses.

People building from inside the market often understand the pain better. They know the cost of a failed payment. They know what happens when crops spoil before reaching town. They know why a clinic visit can take a whole day.

Healthcare Needs Simple, Trustworthy Tools

Healthcare is another area where everyday problems are clear. In many places, clinics are far away. Doctors are limited. Medicine supply can be uneven. Patients may wait too long before getting help. Technology can help here simply with this platform.

A health app must be clear. A pharmacy tool must be reliable. A digital record must protect private data. A remote consultation must connect patients to real care, not just a screen.

Small changes can still have a big effect. Appointment reminders can reduce missed visits. Digital records can help clinics track patients better. Stock tools can show when medicine is running low. Simple phone-based advice can guide people toward care earlier.

The strongest health tech ideas may not look dramatic. They may just save time, reduce mistakes, and help patients avoid long trips.

Transportation Is a Daily Pain Point

Movement affects almost everything. People need to get to work. Goods need to reach markets. Food needs to reach cities. Patients need to reach clinics. When transport is slow or costly, the whole economy feels it.

Many African cities are growing fast. That creates pressure on buses, taxis, motorcycles, delivery vans, and roads. Traffic can waste hours. Poor route planning can raise costs. Informal transport systems can be useful, but hard to track.

A simple route app can help commuters. A delivery platform can help small shops. A fleet tool can help drivers save fuel. A logistics system can help farmers move goods before they spoil.

Local Design Beats Imported Assumptions

One mistake in tech is assuming that one product can work everywhere. Africa is not one market. Even inside one country, users may have different languages, phone types, income levels, network access, and payment habits.

A product built for Lagos may need changes before it works in rural Nigeria. A tool that works in Nairobi may not fit a small town in Ghana. A health system that works in a private clinic may not fit a crowded public facility. Local design matters because details matter.

Can the app work on a low-cost phone? Can people use it with poor internet? Does it support local payment habits? Does it use simple language? Can a person trust it after one bad network day? These questions decide whether a tool becomes part of daily life or gets deleted after one try.

Growth Will Come From Boring Fixes

Some of Africa’s strongest tech growth may come from ideas that sound boring at first. Better invoices. Better delivery tracking. Better crop pricing. Better clinic records. Better payment links. Better transport routes.

Boring tools often solve expensive problems. A shop owner does not care if a product sounds exciting. She cares if it saves money. A farmer does not care if the app wins awards. He cares if it helps sell crops before they spoil. Useful beats flashy.