DD
MM
YYYY

PAGES

DD
MM
YYYY

spot_img

PAGES

Home Blog Page 23

Anthropic Prioritizes Government Issued ID for Claude for New Signups and Subscriptions 

0

Anthropic has started rolling out government-issued ID verification for Claude, including for some new signups and subscriptions. This is a recent change handled through a third-party service called Persona Identities.

A physical, original government-issued photo ID; passport, driver’s license, state/provincial ID, or national identity card from most countries. It must be undamaged, legible, and show your photo. Often required: A live selfie taken in real-time via phone or webcam not a photo of a photo.

Not accepted are Photocopies, scans, screenshots, mobile and digital IDs, student IDs, or temporary paper documents. The process usually takes under 5 minutes. Anthropic states the data is used only for identity confirmation to prevent abuse, enforce policies like the 18+ age requirement, and comply with legal obligations, not for model training. They chose Persona for its privacy and security features.

It’s not yet universal for every basic signup or user—it’s described as a gradual, selective rollout for certain use cases, platform integrity checks, accessing specific capabilities, or subscriptions. Some users including reports of teens or those flagged by age classifiers are hitting it immediately, while others haven’t. However, multiple reports indicate it’s now appearing during new account setup or paid plan activation for many people.

Anthropic has long had an 18+ age policy for Claude and uses classifiers to flag potential underage use, sometimes leading to suspensions. This ID step strengthens enforcement against abuse, fake accounts, and policy violations—without an explicit new regulatory mandate mentioned. It’s the first major consumer AI chatbot like Claude, ChatGPT, Gemini, etc. to implement this level of KYC-style verification.

The move has sparked significant backlash online: Privacy concerns from users who previously saw Claude/Anthropic as more privacy-friendly compared to competitors. Complaints that it creates friction for signups, competitors like ChatGPT still allow quick email, Google and Apple signup without ID. Some users calling it a gift to rivals, as it may drive people away.

Separate issues with false positives on age. This fits a broader industry trend toward stronger identity checks as AI capabilities grow and regulators and companies worry about misuse, but Anthropic is moving faster here than OpenAI or Google on the consumer side. If you’re trying to sign up and hit the prompt, you’ll need a qualifying physical ID handy. For existing users, it may only trigger for certain features or if your account raises flags.

User friction and churn: Adds significant signup and upgrade friction; physical ID + live selfie via Persona. Many users report abandoning subscriptions or upgrades immediately. Reddit and social reactions show strong resistance, with comments like cancelling as soon as this hits and not upgrading anymore.

Widely viewed as handing a gift to rivals. ChatGPT and Gemini still allow quick email and Google signups without ID, while Claude now feels more restrictive. Users and media note this could drive people back to OpenAI/Google, especially after Claude’s recent growth surge. Heavy criticism over sharing sensitive data with third-party Persona.

Users worry about data risks, government ties, and erosion of Anthropic’s privacy-friendly image. Some call it the start of broader AI KYC. Not universal yet—triggers for new accounts, subscriptions, high-usage, or flagged activity. Helps Anthropic enforce 18+ policy and reduce fake and banned accounts, but causes false positives and lockouts for legitimate users.

First major consumer AI chatbot to implement this level of verification. Aims to curb misuse and prepare for regulations, but risks slowing casual adoption and pushing privacy-conscious users toward open-source and local alternatives. Short-term negative sentiment dominates public reaction, potentially slowing Claude’s momentum despite its prior gains in users and app rankings. Long-term depends on how selectively Anthropic applies it and whether competitors follow.

Spartans Casino Hits Top 14 Globally While Bitcoin & Polygon Face a Brutal Slump

0

Both Bitcoin and Polygon have seen sharp drops over the past 90 days, and this shift is changing where people look for gains. Bitcoin has dipped by about 22% in that time, while Polygon has shed over 40% of its value. For two such massive names, that is a very weak showing. It proves that having a big name and past fame is not enough when speed and power vanish. This is the type of market where fans stop waiting for slow recovery tales and start hunting for spots where action and growth still live.

Spartans.com is turning into one of the best examples of this change. Even as the main coins lose their grip, Spartans Casino has seen more than $100 million in deposits in only 60 days and risen to be the 14th biggest crypto casino on earth while still in its test phase. It is a clear sign that eyes are moving toward sites, giving a much faster story for gains.

The Last 90 Days Have Shown Weakness for Bitcoin

The mood for the whole crypto space is still set by Bitcoin, which is why its recent path is so important. A fall of roughly 22% over 90 days is not a small dip for the biggest asset in the sector. It shows a market that has found it hard to keep its faith. Usually, Bitcoin is where people look first for strength and a clear path forward. When it begins losing this much ground over three months, it does more than just hurt its own chart. It breaks the trust of the wider market.

Those holding the coin have not had much reason to feel happy lately. Instead of leading a new jump higher, Bitcoin has been a reason why the market feels sluggish and less sure. For fans searching for better gains, this causes a lot of annoyance.

Polygon Sees an Even Steeper Fall

Recent acts from Polygon have been even weaker than Bitcoin’s, serving as a loud example of how known names can stop moving when market power dies. Over the past 90 days, Polygon has lost more than 40% of its total worth. That is not a dip that fans can just call normal market noise. It shows a real failure to keep buyers interested, even though Polygon stays one of the most famous tokens in the crypto space.

This is what makes the drop so painful. For people watching the market every day, this type of act changes how they see risk. A known name is no longer enough to stay safe. If a giant token can slide this far and still fail to get back up, then eyes naturally start moving toward spots where the growth tale looks more alive and faster.

Spartans Draws Players Because Growth is Clear

This is exactly where Spartans Casino starts to look much more powerful. Its speed is not based on a hope for a comeback or a future jump. The rise is already right there to see. Spartans has seen more than $100 million in deposits in only 60 days and has climbed to be the 14th biggest crypto casino on the globe while still in its test phase. Those are real figures. They show movement and real market power before the site has even started full global work.

In a market where Bitcoin and Polygon both find it hard to make a good upward story, a site showing growth at this level naturally looks better. Spartans is not waiting for a better market time to look good. It is moving right now. Pay-ins are rising, ranks are getting better, and the site is doing all of this before its full start. Spartans is moving toward its formal global opening on August 1, 2026. This means the site is hitting these totals before the next step of growth even begins. Over $100 million in deposits in 60 days is already a giant figure. In a test phase, it means even more.

A big pay-in total only matters if the site behind it is deep enough to keep people interested for a long time. Spartans has that heart. It already gives more than 5,963 games and joins casino and sports play under a single sign-in. It also has crash games, fast games, live staff rooms, table games, top slots, and support for both crypto and cash.

Fast crypto pay-outs add another real lead. Speed is vital when cash is moving, and Spartans has made that a main part of its pull. The site also has the globe’s first $7 million leaderboard live, with $5 million set for the winner and the other $2 million shared among 500 more people. This is not a normal deal. It is the type of move that turns a site into a major market event.

Final Say

Bitcoin and Polygon have both spent the past 90 days showing fans what a lack of speed looks like. Bitcoin is down roughly 22%. Polygon is down over 40%. Neither name is giving the market the type of energy people usually want from big crypto assets. This is why the move toward Spartans makes total sense.

While the giant coins are falling, Spartans has already passed $100 million in deposits in 60 days and rose to 14th on the globe while still in its test phase. Those are not just claims of growth. They are hard facts. Right now, that makes Spartans a much better story than big coins that are failing to earn any attention.

Find Out More About Spartans:

Website: https://spartans.com/

Instagram: https://www.instagram.com/spartans/

Twitter/X: https://x.com/SpartansBet

YouTube: https://www.youtube.com/@SpartansBet

Nvidia CEO Explains Why the Company Invests in All AI Companies, Not Just Front-Runners

0

At a time when the artificial intelligence race is increasingly being framed as a contest between a handful of dominant model makers, Nvidia chief executive Jensen Huang is making a different wager: the future will not be shaped by a single winner, and Nvidia intends to be embedded across the entire field.

Speaking on the Dwarkesh podcast, Huang offered one of the clearest explanations yet of the chipmaker’s unusually expansive investment strategy, one that has seen it back companies across foundation models, cloud infrastructure, robotics, autonomous driving, biotech, and enterprise AI.

“There are so many great, amazing foundation model companies, and we try to invest in all of them,” Huang said. “We don’t pick winners. We need to support everyone.”

That remark has opened a window into how Nvidia sees the next decade of the technology industry. Unlike a traditional venture capital firm seeking outsized returns from a small set of bets, Nvidia is investing with a far broader strategic purpose: ensuring that whichever company emerges stronger in the AI race, it does so on Nvidia’s infrastructure, software ecosystem, and silicon.

In effect, Nvidia is not merely funding startups; it is underwriting demand for its own future. Huang said the reasoning is rooted partly in humility and partly in history.

“When Nvidia first started, there were 60 3D graphics companies,” he said. “If you would have taken those 60 graphics companies and asked yourself which one was going to make it, Nvidia would be at the top of that list not to make it.”

In the early 1990s, Nvidia was one among dozens of companies chasing the graphics market, and few on the outside would have predicted that it would become the world’s most valuable company and the nerve center of the AI boom.

Huang acknowledged as much.

“Everybody would have counted us out,” he said. “And here we are. So I have enough humility to recognize that. Don’t pick winners.”

For seasoned market watchers, this is a disciplined strategic posture shaped by Nvidia’s understanding of platform economics. Every successful AI lab, robotics firm, self-driving company, or biotech platform creates downstream demand for compute. Every model trained, every inference request processed, every AI agent deployed translates into more GPUs, networking equipment, software licenses, and cloud infrastructure.

This is where Nvidia’s investment approach becomes especially significant. The company is reinforcing the entire AI stack, from the model layer to deployment and real-world use cases. Huang has previously described AI as a “five-layer cake,” spanning energy, chips, data centers, models, and applications.

Seen through that lens, Nvidia’s capital deployment is a deliberate effort to maintain influence across every layer of that architecture. Its recent investments illustrate the scale of that ambition. In November, Nvidia committed up to $10 billion to Anthropic, the company behind Claude, while in February it announced a $30 billion investment in OpenAI.

These are not marginal venture checks but strategic balance-sheet moves that underscore how central foundation model companies have become to Nvidia’s long-term outlook. The company has also backed Mistral AI, Europe’s leading frontier lab, further broadening its exposure to the global model race.

Beyond large language models, Nvidia’s investments extend into adjacent sectors that are likely to define the next phase of AI commercialization. These include Wayve in autonomous driving, Scale AI in data labeling and model infrastructure, and Figure AI in humanoid robotics.

This breadth indicates that Nvidia believes the most durable value in AI will not reside solely in chatbots or consumer-facing assistants, but in the industrial and enterprise applications that emerge from the technology.

Robotics, physical AI, autonomous mobility, and AI-enabled scientific discovery are all compute-intensive domains where Nvidia’s chips and software remain foundational.

There is also a more subtle competitive logic at work. By investing broadly rather than narrowly, Nvidia reduces the strategic risk of backing the wrong horse.

Whether the eventual leader is OpenAI, Anthropic, Mistral, or a company not yet on the radar, analysts believe Nvidia stands to benefit so long as these firms continue to build on its hardware and CUDA ecosystem.

That is what makes Huang’s “don’t pick winners” line so consequential. It is less an expression of neutrality than a declaration of platform confidence. Nvidia no longer needs to predict who wins the AI race because it has positioned itself to profit from nearly every serious contender.

TSMC Delivers Another Record Quarter, 58% Surge in Net Profit, as AI Demand Intensifies

0

Taiwan Semiconductor Manufacturing Company posted another blockbuster performance on Thursday, reporting a 58% surge in first-quarter net profit that shattered market expectations and extended its streak of record earnings.

The results underscore the unrelenting appetite for advanced chips powering the artificial intelligence boom, even as fresh geopolitical risks swirl around global supply chains.

Net income for the three months ended March reached NT$572.48 billion, comfortably ahead of the LSEG SmartEstimate of NT$543.32 billion. Revenue climbed to NT$1.134 trillion (about $35 billion), topping the expected NT$1.127 trillion and marking a 35% jump from a year earlier. This marks the fourth consecutive quarter of record profits for the world’s largest contract chipmaker.

TSMC first flagged the strong revenue growth last week, but the profit beat provided fresh confirmation of healthy margins and operational efficiency. Asia’s most valuable technology company by market capitalization, TSMC produces the sophisticated semiconductors found in everything from iPhones to data-center accelerators. Its biggest customer is now Nvidia, whose AI processors have become the engine of the current technology cycle, while Apple remains a key partner for mobile chips.

“AI-related demand continues to be extremely robust,” TSMC President and CEO C.C. Wei said during Thursday’s earnings call.

He pointed to rapid advances in artificial intelligence that are driving exponential increases in computation needs. Wei added that the company has received “strong signals and a positive outlook” from customers, reinforcing its belief in a multi-year AI growth cycle.

The company guided for full-year 2026 revenue growth of more than 30% in U.S. dollar terms and projected second-quarter revenue between $39 billion and $40.2 billion, implying roughly 10% sequential growth. It also signaled that capital spending this year will land at the high end of its previously announced range of $52 billion to $56 billion, reflecting aggressive expansion to keep pace with demand.

Advanced process technologies, defined as 7-nanometer and below, accounted for 74% of wafer revenue in the quarter, with chips at the leading-edge 3-nanometer node alone making up 25%. High-performance computing, which includes AI and 5G applications, drove 61% of total sales, highlighting how deeply the AI wave has reshaped TSMC’s business mix.

Smaller nanometer processes deliver greater transistor density, translating directly into higher performance and better power efficiency—the very attributes hyperscalers crave for training and running ever-larger AI models.

Executives used the call to address potential risks from the ongoing Middle East conflict, including possible disruptions to energy supplies and specialty gases such as helium and hydrogen. They reassured investors that near-term operations face no material impact, thanks to diversified sourcing and safety stockpiles of critical materials.

The company is also expanding physical capacity. It recently broke ground on a new advanced fabrication plant in Tainan, Taiwan, as part of a broader global push that includes facilities in the United States and Japan.

Senior analyst William Li at Counterpoint Research captured the current market dynamic, noting: “The narrative for 2026 is as much about resource constraints as it is about growth. Demand still significantly outpaces supply and isn’t showing any major sign of slowing down.”

Li added that the industry is likely to remain in a sold-out environment throughout the year, with semiconductor companies struggling to keep products on shelves. That tightness helps explain TSMC’s decision to lean into the upper end of its capex guidance and why shares have continued to perform well despite broader market jitters.

For all the optimism, the scale of investment required is staggering. TSMC is pouring tens of billions into new fabs and equipment at a time when energy prices remain elevated, and geopolitical tensions from the Strait of Hormuz to U.S.-China tech frictions add layers of uncertainty.

Yet Wei and his team appear convinced that the AI megatrend is structural rather than cyclical, driven by applications that are only beginning to move from data centers into everyday use.

The latest results leave little doubt that TSMC sits at the center of the AI supply chain. As long as hyperscalers, smartphone makers, and emerging AI-adjacent sectors keep placing orders for its most advanced nodes, the foundry giant is positioned to keep printing record numbers. The bigger question for 2026 and beyond is whether the industry can expand capacity fast enough to meet that insatiable demand without triggering the kind of shortages or cost spikes that could eventually cool the boom.

For now, TSMC’s numbers suggest the party is far from over.

OpenAI Announces the Release of GPT-5.4-Cyber Variant of GPT-5.4- Model

0

OpenAI has announced the release of GPT-5.4-Cyber. It’s a specialized variant of their flagship GPT-5.4 model, fine-tuned specifically for defensive cybersecurity use cases. It lowers the refusal boundary for legitimate cybersecurity tasks compared to the standard GPT-5.4.

This makes it more willing to assist with potentially sensitive or risky prompts like analyzing code for vulnerabilities when used by verified defenders, while still maintaining safeguards against clear malicious intent. New capabilities includes enhanced support for binary reverse engineering. Security professionals can use it to analyze compiled software (binaries) for: Potential vulnerabilities, Malware indicators, Overall security robustness— without needing access to the original source code.

Access model is not publicly available. It’s rolling out through OpenAI’s expanded Trusted Access for Cyber (TAC) program, initially limited to vetted and high-tier customers such as security vendors, organizations, researchers, and authenticated cyber defenders. Interested parties in the program can apply for additional access tiers.

This follows similar moves by competitors like Anthropic’s recent Mythos model focused on cyber. OpenAI positions it as part of scaling defenses in lockstep with advancing AI capabilities, including prior efforts like their Cybersecurity Grant Program and Codex Security tools that have reportedly helped fix thousands of vulnerabilities in open-source projects.

AI models are increasingly capable in areas like code analysis and autonomous reasoning, which has a natural dual-use risk, helpful for defenders but potentially dangerous if misused by attackers. GPT-5.4-Cyber represents OpenAI’s approach: make powerful tools more available to the good guys (under controlled access) while iteratively improving safety.

It’s part of their broader Preparedness Framework for high cyber-risk capabilities. The model lowers refusal boundaries for legitimate tasks, enabling faster vulnerability discovery, malware analysis, and binary reverse engineering; analyzing compiled software without source code. This helps defenders identify and fix issues quicker than attackers.

OpenAI is scaling its Trusted Access for Cyber (TAC) program to thousands of verified individual defenders and hundreds of teams protecting critical software. Highest-tier users get the more permissive GPT-5.4-Cyber variant. Access remains limited and vetted via ID checks and authentication to minimize misuse.

Competitive response: Released just days after Anthropic’s Claude Mythos, it positions OpenAI as favoring broader yet controlled defender access versus more tightly gated approaches. It builds on prior efforts that reportedly helped fix thousands of vulnerabilities in open-source projects.

By making the model cyber-permissive only for vetted users, OpenAI aims to give defenders an edge while testing for jailbreaks and improving safeguards. It acknowledges growing AI use by attackers but argues current controls sufficiently reduce immediate cyber risks. Security vendors like CrowdStrike integration mentioned in early reactions and researchers gain advanced tools for proactive defense, potentially improving threat detection and secure coding at scale.

Broader rollout beyond initial testing could follow based on feedback. Not available publicly or to general ChatGPT users. U.S. government agencies are not yet included (discussions ongoing). Focus remains on iterative, responsible deployment to stay ahead of adversarial use. The move emphasizes scaling defenses in lockstep with AI capabilities, prioritizing verified “good guys” while monitoring risks. Early expert reactions view it positively for empowering defenders, though success depends on effective vetting and ongoing safety improvements.

If you’re a cybersecurity professional, you can check OpenAI’s official blog post for details on applying to the Trusted Access program. The announcement emphasizes empowering defenders to find and fix issues faster than adversaries can exploit them.