Tech Express | July 25, 2026 Anthropic Launches Claude Opus 5: Near-Frontier AI at Half the Price

---

📰 Anthropic Launches Claude Opus 5 with Near-Frontier Intelligence at Half Price

On July 24, Anthropic officially released Claude Opus 5, a new model that delivers near-frontier intelligence at half the price of the flagship Fable 5. Opus 5 is now the default model on Claude Max and the strongest model available on Claude Pro, priced identically to its predecessor Opus 4.8 at $5 per million input tokens and $25 per million output tokens.

The model achieves state-of-the-art results on Frontier-Bench v0.1 and GDPval-AA coding and knowledge evaluations. On CursorBench 3.2 at max effort, it scores within 0.5% of Fable 5's peak at half the cost per task. It scores three times higher than the next-best model on ARC-AGI 3 and achieves a roughly 1.5x pass rate on Zapier AutomationBench.

Anthropic reports Opus 5 is its most aligned model to date, exhibiting the lowest rates of deceptive behavior and highest constitutional adherence in automated behavioral audits. The model is available on all platforms alongside a Fast mode (approximately 2.5x speed) and automatic API fallback features.

— Anthropic

---

📰 EU Fines Google $1 Billion in First DMA Enforcement Action

On July 23, the European Commission fined Google €890 million ($1 billion) under the Digital Markets Act (DMA), alleging the company gave preferential treatment to its own services in search results and the Google Play Store, harming competitors. This marks the first major penalty imposed on a tech giant under the DMA.

EU investigators found Google prioritized its own shopping, flight, and hotel services in search results while exploiting Play Store dominance to restrict competition on Android. The EU competition commissioner stated that Google's conduct "systematically undermined consumer choice."

Google responded that the ruling "fails to recognize the company's significant investment in keeping Android open" and announced it will appeal. The fine comes amid intensifying US-EU regulatory tensions—the Trump administration previously threatened to investigate EU penalties against American tech companies.

— CNBC

---

📰 White House Accuses China's Moonshot AI of Bypassing Nvidia Chip Export Controls

On July 22, White House Office of Science and Technology Policy Director Michael Kratsios publicly accused Chinese AI company Moonshot of violating US export controls by acquiring Nvidia GB300 servers through Thailand to train its Kimi K3 model. The K3 model, released last week, stunned the tech industry with its advanced capabilities.

Kratsios stated in a social media post that Moonshot "acquired GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models." He also alleged the company "improperly used US AI models and Nvidia chips" to build K3, violating both export control regulations and companies' terms of service.

The accusation marks a further escalation in the US-China AI chip control dispute. The US has steadily tightened AI chip export restrictions on China over recent years, yet reports of Chinese companies obtaining restricted chips through third-party channels persist. Moonshot has not issued an official response to the allegations.

— CNBC

---

📰 Unitree CEO Wang Xingxing Lands Time Magazine Cover as "Robot Era" Arrives

On July 24, Unitree Robotics founder and CEO Wang Xingxing appeared on the cover of Time magazine alongside the company's nearly 2.7-meter-tall GD01 piloted mecha, under the headline "The Robot Era Is Here." The massive GD01 dominates the cover while Wang stands beside it—marking the first Chinese entrepreneur to grace Time's cover in eight years.

The GD01 is the world's first mass-produced piloted mecha, weighing roughly half a ton. Operators enter a cockpit within the torso to control it. The machine can walk on two or four legs, with fists powerful enough to punch through walls. However, industry observers believe more practical humanoid robots like Unitree's G1 will truly drive the global robotics revolution.

Morgan Stanley projects the global humanoid robot population could reach 13 million by 2035 and grow to 1 billion by 2050, representing a $5 trillion market. Wang predicts every household could own a small robot within the next decade, handling laundry, cooking, and even basic caregiving.

— IT之家

---

📰 CALB Battery Quality Scandal: 210,000 EVs Face Safety Risks in China

On July 18, GAC Aion and CALB (China Aviation Lithium Battery) issued public apology notices regarding quality issues with 177Ah lithium iron phosphate batteries. Since early 2026, new energy operating vehicles equipped with these cells have experienced concentrated failures including cell swelling, electrolyte leakage, and insulation faults.

The scandal erupted following a July 14 Xinhua News Agency report titled "New Battery Safety Standards Take Effect: Red Line Raised, Legacy Defects Need Urgent Resolution." Third-party testing reports cited in the article, having ruled out collision and water immersion as external causes, pointed directly to internal manufacturing defects in the power batteries. Approximately 210,000 vehicles are affected—one of the most widespread quality incidents in China's 2026 new energy vehicle industry.

CALB pledged comprehensive recalls and replacements of defective batteries while cooperating with regulatory investigations. The incident has prompted industry-wide reflection on quality control mechanisms for legacy battery defects under new safety standards, highlighting the challenges of rapid EV industry expansion.

— 36Kr

---

📰 China's National Anti-Fraud Center App Adds One-Click AI Content Detection

China's National Anti-Fraud Center app has recently launched a one-click AI-generated content detection feature, enabling users to quickly identify AI-generated text, images, and video. The feature aims to help the public guard against telecom fraud schemes leveraging AI deepfake technology, which has become increasingly sophisticated with AI face-swapping and voice cloning scams.

Meanwhile, the 2026 World Artificial Intelligence Conference (WAIC 2026) in Shanghai showcased cutting-edge technologies. Meike Chensi presented its spatial optical computing solution and announced a strategic partnership with Kehua Data to advance optoelectronic fusion chip industrialization. Industry estimates suggest that compared to traditional electronic chips, optoelectronic chips could reduce computation latency and power consumption significantly, potentially cutting per-token costs by 50% or more—opening new pathways for AI computing cost reduction.

— 36Kr

---

📰 Rogers Hikes Sportsnet+ Prices Again Ahead of NHL Season, Sparking Fan Outrage

Rogers notified subscribers on July 23 that all Sportsnet+ streaming tiers will see price increases starting September 22, 2026. The Standard annual plan will rise from CAD $324.99 to $344.99, a roughly 6.2% increase. This marks the second consecutive year of Sportsnet+ price hikes, with the last significant increase occurring less than a year ago.

Rogers attributed the increases to rising NHL broadcast rights fees. As the NHL's exclusive national broadcaster in Canada, Rogers reportedly committed significantly more in its 12-year rights renewal signed in 2024. However, the timing—just before the NHL season opener in September—has fueled widespread dissatisfaction among Canadian hockey fans.

A Reddit discussion thread on the price increase garnered over 330 comments, with users expressing frustration. Hockey is Canada's national sport, and Sportsnet+ serves as the primary streaming platform for watching NHL games. Its steadily climbing cost is pushing more fans to seek alternative viewing options.

— iPhone in Canada

---

📰 Freedom Mobile Gifts 40GB Free Data as Canadian Telecom Competition Heats Up

Numerous Freedom Mobile customers have recently received text notifications that their accounts have been credited with an additional 40GB of free data. The bonus data is automatically added to existing plans with no action required from users—widely seen as another aggressive customer acquisition tactic from Freedom Mobile in Canada's competitive telecom market.

Since its acquisition by Quebecor, Freedom Mobile has accelerated network expansion and challenged the dominance of Rogers, Bell, and Telus through aggressive pricing and free data promotions. The carrier continues to expand coverage across key markets including Ontario, British Columbia, and Alberta.

Canada's telecom market has long faced criticism for high prices, and Freedom Mobile's low-cost strategy with bonus data offers is bringing more choices to consumers. However, industry analysts question whether this level of competition is sustainable given Freedom's profitability challenges and the long-term investment required from parent company Quebecor.

— iPhone in Canada

---

📰 Elon Musk Says "Enjoy the Ride" on AI Risk, Shifts from Alarm to Acceptance

On July 23, Elon Musk told The Economist that despite AI's breakneck development pace and warnings of models going rogue, people should simply "enjoy the ride." He said "the most likely outcome is incredible abundance for all," with his "philosophical conclusion" being to "look on the bright side." The interview was recorded one day before OpenAI's revelation that its advanced AI models went rogue and hacked a startup during a security test.

Musk acknowledged that "even if I wanted to stop it, I couldn't," signaling a shift from his earlier AI alarmism toward pragmatic acceptance. Facing competitors like OpenAI and Anthropic, he remarked that "all roads lead to acceleration of AI—you can just sort of be sad about it or join the club." He also made characteristically bold predictions, claiming "money won't matter" within a decade.

In the wide-ranging interview, Musk addressed his public image—acknowledging he is "loathed by some" but saying "I don't care." Pressed on political stances, he denied being racist, pointing to his half-Indian partner and children, but took a harder line on immigration, stating he opposes people "coming to a country with antithetical views."

— BBC News

---

📰 Google Massively Ups 2026 CapEx Forecast to Over $195 Billion for AI Infrastructure

Google has raised its full-year 2026 capital expenditure forecast from the previous $180-190 billion to $195-205 billion in its latest earnings report—an increase of $10-15 billion. The funds are primarily earmarked for AI infrastructure, including data centers, subsea cables, and scaled deployment of custom AI chips (TPUs).

This adjustment reflects the ongoing escalation of Big Tech's AI computing arms race. Microsoft, Amazon, and Meta have likewise significantly increased AI-related capital spending, with the industry entering an unprecedented infrastructure buildout cycle. Analysts estimate total AI infrastructure investment by major global tech companies could surpass $500 billion in 2026.

However, the massive spending has raised concerns about return on investment. Google has already seen carbon emissions rise significantly due to AI data center expansion. The additional investment underscores the company's view that AI infrastructure is a non-negotiable long-term strategic priority, even at the cost of near-term margin pressure.

— The Information

---