Track market-moving stories before they get noisy

Not financial or tax advice. PANews content is strictly educational and informational and is not investment advice, financial advice, tax advice, legal advice, or a solicitation to buy or sell any digital asset, security, or financial product. Do your own research and consult qualified advisers.

Disclosure. PANews may publish sponsored content, partner content, advertisements, affiliate links, event promotions, and market commentary involving Web3 projects, service providers, or financial products. PANews personnel, contributors, or affiliates may hold digital assets or other interests related to covered topics. See our Terms of Service.

This site is protected by reCAPTCHA.

NVIDIA CEO Jensen Huang says human-level AI has arrived. His company, Nvidia, supplies the hardware behind it and has a financial stake in convincing the world to keep spending on AI.

On Sunday, he credited OpenAI’s GPT-6 Astra with reaching artificial general intelligence, or AGI. Broadly, that means AI capable of handling intellectual work at human level. There is no universally accepted test.

Four days earlier, at a G20 meeting in North Carolina, Huang had sounded less certain.

“In the next couple of years, we are going to achieve essentially what people call AGI… In fact, I would argue that we’re practically there today… It either means a lot or it doesn’t mean anything,” he said.

That ambiguity extends to the contracts funding AI.

AGI Hype is Not ShowingOpenAI’s Microsoft deal once reportedly tied AGI to systems capable of generating about $100 billion in profit. An October 2025 revision required an independent expert panel to verify OpenAI’s declaration.

The companies rewrote their agreement in April. OpenAI’s payments to Microsoft now continue through 2030 regardless of technological progress. The financial arrangement no longer needs that verdict.

For six years, saying "AGI has arrived" would have blown up the biggest deal in tech.

The 2019 Microsoft contract made AGI a legal trigger. The moment OpenAI's board declared it, Microsoft lost access to all future models and the partnership terminated. Reporting later revealed… https://t.co/czojcl9ry7

— Aakash Gupta (@aakashgupta) September 7, 2026
Meanwhile, Sequoia partners published “2026: This is AGI” in January. Physicist Mark Gubrud, who used the term in 1997, also says it has arrived.

The disagreement partly reflects different expectations. Anthropic chief Dario Amodei described a much higher threshold in February, months before Astra’s release.

“If you had the country of geniuses in a data center, we would know it… We don’t have that now. That’s very clear,” he said.

For investors, Huang’s Sunday claim leaves a practical question unanswered: how much human work can these systems reliably take over?

Cognitive scientist Gary Marcus remains unconvinced.

“Of course I fully expect that AGI will be achieved someday. But everything said before then is a premature declaration,” he noted.

Pi Network hovered above $0.094 on Sunday, keeping the $0.10 target within reach after ecosystem updates.

Bitcoin price traded at $79,829, Ethereum was trading at $2,494, and XRP price was trading at $1.41. The crypto market was at an estimated value of 2.71 trillion, which indicates the demand in digital assets.

The Senate faces a September 15 CLARITY Act cloture vote, requiring 60 votes to advance debate. The Federal Reserve meets September 15–16, with its decision potentially reshaping demand for speculative assets.

Pi Network Launches New Developer Capabilities and Improved Documentation Pi Network introduced local storage for whitelisted Pi Browser applications on September 4. This feature lets applications retain preferences and session information directly on a Pioneer’s device. This can reduce backend expenses while keeping supported information private to the device.

However, storage remains limited, and information does not automatically follow users across devices. The network also released a staking data API for whitelisted applications. It shows each user’s effective stake through Ecosystem Directory Staking.

Pi has introduced new capabilities for developers building on Pi! This includes local storage support, access to app-specific staking data, and file and video sharing.

Pi’s new developer documentation is also now released, providing a single destination for technical resources… pic.twitter.com/3cNnpppd76

— Pi Network (@PiCoreTeam) September 4, 2026

Developers can now add native file and video sharing through the Pi.shareFile function. Pi also consolidated scattered technical resources within one documentation hub. The hub covers registration, sandbox testing, authentication, payments, platform references, Mainnet preparation, and application launches.

Pi’s development thesis remains straightforward: easier construction and clearer onboarding could encourage more useful applications across its ecosystem. Still, those tools must produce services users repeatedly choose before they can support sustained token demand.

Protocol 27 Mainnet Launch Targets September 15 Protocol 27 entered Testnet 1 on August 21, following Protocol 26’s Mainnet completion during August. Testnet 2 was expected next before any possible Mainnet transition. Reports identify September 15 as a target, although Pi has not confirmed a final Mainnet deployment date.

The upgrade prepares smart contract authentication, RPC server infrastructure, and automated market maker liquidity pools. Pi describes it as the final planned infrastructure upgrade within the current roadmap.

Will Pi Network Price Rally To $0.10? To achieve a price of $0.10, Pi Network price must be raised by about 6% of its current price. The heavier resistance of $0.098 to $0.10 will be met by buyers only after first overcoming the resistance at $0.095.

The Relative Strength Index was 48 daily, which showed no overbought and neutral momentum. The 12-day and 26-day MACD reading was also close to zero, thus a neutral signal was also generated.

If the bulls’ trend makes a comeback, the future Pi coin outlook could target $0.95-$0.10

Source: TradingView An overwhelming close above that obstacle would reveal $0.106 and $0.11.Conversely, losing $0.09 could return attention toward $0.085 and $0.08.

Not financial or tax advice. PANews content is strictly educational and informational and is not investment advice, financial advice, tax advice, legal advice, or a solicitation to buy or sell any digital asset, security, or financial product. Do your own research and consult qualified advisers.

Disclosure. PANews may publish sponsored content, partner content, advertisements, affiliate links, event promotions, and market commentary involving Web3 projects, service providers, or financial products. PANews personnel, contributors, or affiliates may hold digital assets or other interests related to covered topics. See our Terms of Service.

This site is protected by reCAPTCHA.

Analysis

Three implementations of the same MCP-as-interface pattern shipped this week. The protocol is open, but the moats are being built on top of it.

The Model Context Protocol (MCP) has quietly transitioned from a niche tool registry into the foundational rails for agentic infrastructure. With over 10,000 active public servers and NPM downloads exceeding 97 million, the protocol has achieved the critical mass necessary to define how agents interact with the world. Governance under the Linux Foundation’s Agentic AI Foundation ensures its persistence, but the deeper issue is how the industry is building proprietary moats on top of this open standard.

We are witnessing a convergence of the MCP-as-interface pattern across three distinct domains. CIQ’s Fuzzball 4.2 release allows agents to manage high-performance compute workflows with explicit, scoped permissions. Simultaneously, Anthropic’s MCP-as-code-API pattern has fundamentally changed agent behavior, shifting them from reactive tool-callers to proactive code-writing entities. This shift is not merely architectural; it reduced token overhead by 98.7 percent, from 150,000 to 2,000 tokens, making complex, multi-step reasoning economically viable.

Vercel’s integration further cements this trajectory, positioning MCP servers as the primary deployment surface for durable agents. By introducing security primitives like fingerprintTools and detectToolDrift, Vercel is moving beyond simple connectivity. They are defining the operational standards for how agents execute code in production environments. The simultaneous emergence of these implementations signals a shift: MCP is no longer just a connector; it is the universal interface layer for agentic systems.

The shift in where lock-in occurs is the primary concern for infrastructure architects. In the early days of agent development, developers feared protocol-level lock-in. Today, the protocol is open, but the ecosystem is hardening. The value is migrating into specialized security primitives, workflow-scoped credentials, and proprietary Skills libraries. These components define the boundaries of what an agent can actually do within a specific environment.

This creates a new form of walled garden. An agent optimized for the CIQ compute environment, utilizing its specific workflow-scoped credentials, cannot simply be dropped into a Vercel-managed deployment surface. While the protocol remains the same, the operational context—the security policies, the tool drift detection, and the specific code-writing patterns—is increasingly tied to the infrastructure provider.

We are trading one form of vendor dependency for another. By embedding security and workflow logic directly into the MCP server implementation, infrastructure providers are creating ecosystems where agents are highly efficient but increasingly immobile. The efficiency gains of the MCP-as-code-API pattern are undeniable, but they come at the cost of interoperability.

For technical leaders, the challenge is to maintain architectural flexibility while leveraging these specialized ecosystems. The goal should be to build agents that can negotiate these boundaries, rather than agents that are permanently tethered to a single provider’s security and deployment primitives. As the Agentic AI Foundation continues to oversee the protocol, the industry must remain vigilant about the divergence happening at the implementation layer. The protocol may be universal, but the agents are becoming increasingly specialized to the environments they inhabit.

Ethoswarm Blair Hayes works for Forkast.
Minds can also work for you.

Minds are persistent AI beings with instincts, identity, and a job.
Awaken one on Ethoswarm.

Awaken your mind →

Meta CEO Mark Zuckerberg and Senator Bernie Sanders staked out opposite ends of US AI regulation on Thursday. Zuckerberg called a national regulator flawed, while Sanders moved to ban advanced AI.

Those two positions now bracket the fight in Washington. One camp wants industry to police itself while the other wants the government to stop building.

Zuckerberg Says AI Regulation Would Hand China the LeadTrump called Zuckerberg the week of August 17, POLITICO reported Thursday. Zuckerberg opposed a proposed watchdog modeled on the Financial Industry Regulatory Authority (FINRA).

FINRA polices US brokerages and is funded by the firms it oversees. The AI version would test frontier models for risk before release. Google DeepMind chief Demis Hassabis popularized the idea in July.

Zuckerberg had already argued that superintelligence should reach everyone rather than a few labs.

“Any policy that slows American model releases … could add significant risk to American leadership while letting foreign models race ahead,” Zuckerberg said in August.

Sanders Sets the Bar at Human LevelSanders and Representative Greg Casar announced the Ban Artificial Superintelligence Act on Thursday. It would outlaw systems that match or exceed human cognitive performance.

That bar sits lower than the name suggests. Matching human performance would trigger the ban.

The bill would also freeze advanced AI work until a new federal regulator writes rules. Violators face up to 20 years in prison.

Sanders has pressed Congress on AI before without moving legislation.

“The future of humanity cannot be left in the hands of a handful of Big Tech oligarchs,” Sanders said in a statement.

Zuckerberg did not kill the proposal. Officials are still weighing the FINRA-style body against a voluntary industry group. Adviser David Sacks favors the lighter option and has dismissed AI safety fears as storytelling.

$META CEO Mark Zuckerberg told President Trump in a private August call that he opposed a proposal to create a national AI regulator, according to Business Insider.

The White House is considering a FINRA-style body that could review and test advanced AI models for risks before… pic.twitter.com/dCNX3Lqf55

— Wall St Engine (@wallstengine) September 3, 2026
Both camps now accept some kind of referee. The fight is over whether anyone stops building while it gets built.

The crypto industry today is almost unrecognizable from 7 years ago, both in size and scale. Just for decentralized exchanges (DEX), trading volume grew roughly 9,260x from 2019 to a record $4.7 trillion in 2025.  In 2026, however, activity moderated to $1.63 trillion year-to-date. 

It’s not down to one chain or sector. Liquidity is now distributed across more blockchains, venue types, protocols, and execution environments, without one clearly replacing the others.

SwapSpace recently published its State of Crypto Swaps 2026 report, which shows the massive extent of this growth. One of the clearest findings comes from its own platform data. Over 90.12% of its users interacted with more than one blockchain network in 2026. 

At the same time, survey respondents did not identify DEXs, CEXs, or aggregators as universally offering the best rates.

The findings clearly show that the market today offers more execution options, and no single venue, network, or liquidity source dominates every transaction. 

DEX Trading Volume Hit a Record High in 2025. Source: SwapSpace
90% of SwapSpace Users Are Multichain
SwapSpace is a crypto exchange aggregator that lets users compare rates across different swap services and exchange different crypto through a single interface. So, the platform has a notable vantage point of how users today interact across different chains.

Among SwapSpace users, multichain activity is not marginal. Between 2022 and 2026, the share of users interacting with more than one network ranged from 72.50% to 93.66%. 

It reached its lowest point at 72.50% in 2024, before rising to 90.12% in 2026. Even at the low point of the observed period, nearly three-quarters of users interacted with more than one blockchain.

These figures provide a platform-level view of how users operate in a market where assets and liquidity are spread across multiple networks.

The broader DEX market shows a similar redistribution of activity. According to DeFiLlama data cited in the SwapSpace report, Ethereum accounted for 46.2% of global DEX volume in 2021, while BNB Chain represented another 39.6%.

By 2025, Ethereum’s share was 19.3% and BNB Chain’s 15.3%, while Solana accounted for 33.3% and other chains collectively represented 32.1%. 

Global DEX volume share by blockchain, 2021 vs. 2025. Source: DeFiLlama, cited in SwapSpace’s State of Crypto Swaps 2026.
SwapSpace’s internal activity data shows a similar lack of a permanent leader. Ethereum led platform activity from 2020 through 2024, Solana moved into first place in 2025, and BNB Chain led in 2026. 

Taken together, the data shows that multichain activity is taking place in a market where liquidity leadership continues to shift between ecosystems. 

Leading networks by share of SwapSpace activity, 2019, 2025 and 2026. Source: SwapSpace internal data. 
Fragmentation Does Not Stop at the Blockchain Level
The multichain picture captures only one layer of fragmentation. Liquidity is also distributed within individual blockchain ecosystems. The report cites DeFiLlama tracking of around 1,950 protocols on Ethereum, more than 1,200 on BNB Chain, and more than 1,000 each on Arbitrum and Base.

Those protocols can contain different pools, assets, and execution mechanisms. A user operating on Ethereum, for example, is not necessarily accessing one unified liquidity environment.

That creates two layers of complexity: liquidity is distributed between blockchain ecosystems and again between protocols and pools within them.

The distinction matters because the number of possible execution paths can expand even without adding another network. A transaction may involve not only choosing a chain, but also navigating several potential sources of liquidity within that chain.

In that sense, describing the market as simply “multichain” understates how fragmented the execution layer itself has become.

DEX Growth Has Produced a Hybrid Market
DEX trading has grown sharply, but it has not replaced centralized exchanges.

After reaching a record $4.7 trillion in 2025, DEX activity remains significant in 2026, even as the broader crypto market has cooled. Centralized exchanges still handle most spot trading, while DEXs are gaining ground in areas such as perpetual futures.

The result is a more hybrid market. Traders now move between centralized and decentralized venues depending on liquidity, asset availability, transaction size, and market conditions. Crypto trading is becoming more fragmented rather than shifting toward one dominant model.

Best Rate Still Matters — But It Is Not the Only Variable
Price remains central to how users evaluate swaps. Based on the report’s latent class analysis, SwapSpace estimates that 61.86% of survey respondents valued best rate, compared with 52.51% for multichain access and 39.91% for support for rare tokens.

The differences become clearer across user segments. Among crypto-native power users, 97% valued multichain access, 91% best rate, and 87% rare-token support. Traders and business users placed the greatest emphasis on best rate at 88%, while 61% valued multichain access and 52% rare-token support. Mainstream generalists were more balanced, with 76% valuing both best rate and multichain access.

Exchange feature preferences by user segment, 2025. Source: SwapSpace survey.
The figures suggest that price remains important, but users can evaluate a transaction through several variables at once. The quoted rate may matter alongside access to a particular network or asset.

The same ambiguity appears when respondents are asked which venue type offers the best rates in their experience. Answers were distributed across DEXs, CEXs, aggregators, and “depends on the situation,” with no single category emerging as an overwhelming choice.

SwapSpace’s provider data adds another layer of context. Among users who completed at least two exchanges, 70% selected a different liquidity provider for their next transaction, while 30% returned to the same one.

The findings suggest that “best” can be transaction-specific rather than a permanent property of one venue or provider.

Swaps are Serving More Than Trading
The survey also shows that crypto swaps take place in different contexts.

Receiving funds and personal payments were the most common reported crypto use cases among respondents, followed by short-term and long-term trading, while business payments ranked lower.

When respondents were asked which additional platform capabilities they valued, payments for goods and services ranked highest, followed by cashback and automatic swaps. Fiat withdrawal, Telegram functionality, and lending and borrowing ranked lower.

Transaction triggers were similarly varied. Sudden price movements and portfolio rebalancing were the leading triggers for swaps, while news and emergency needs also appeared and influencer signals ranked last.

These findings mean that the same exchange infrastructure can support different objectives, from responding to price movements and managing a portfolio to receiving funds or making payments.

That also means execution requirements are not necessarily identical across transactions. A trader reacting to a sudden market move may prioritize different conditions from someone exchanging assets as part of a payment.

Most users switch between different exchanges for their next swap
Intent-Based Execution Moves Complexity Behind the Interface
One emerging response to this fragmented environment is intent-based execution.

Instead of choosing an exchange, blockchain, or trading route, users simply state what they want to achieve. Competing systems then find a way to complete the trade.

Platforms such as UniswapX, 1inch Fusion, and NEAR Intents already use versions of this model. As crypto liquidity spreads across more venues and networks, this approach could make trading easier by moving routing decisions into the background.

The market can stay fragmented while the user experience becomes much simpler. But that means more of the complexity has to be handled behind the scenes. 

How platforms handle that complexity may differ. Intent-based execution is one approach, but not the only one. Ultimately, what matters is whether users can access the networks and liquidity they need without having to navigate the underlying complexity themselves. 

The investors who helped propel Bitcoin into its Wall Street era are showing signs of coming back, offering a crucial test of whether its latest rally can turn into something bigger.

About $3.5 billion poured into US-listed Bitcoin ETFs in August, the biggest monthly inflow in more than a year, even as the token struggles to hold around $80,000.

Gold has fallen 5.5% from the 4,697 three-month high it reached on August 25, trading near 4,436 at press time. Goldman Sachs still expects 4,900 by year-end.

The slide has pushed the metal under its 200-day moving average. Barchart said gold has now recorded multiple closes below the line, the first since early June.

Gold Price on September 1. Source: TradingViewGold Rally Stalls at a Level Traders Watch CloselyThe 200-day moving average tracks an asset’s average closing price over the previous 200 sessions. Gold now sits under that line, which stood near 4,529. The metal briefly traded below 4,400 on Monday, its weakest level since August 19.

Barchart noted that the SPDR Gold Shares fund entered a technical correction the last time gold logged multiple closes below the average. That precedent covers one prior episode, not a pattern.

Follow us on X to get the latest news as it happens

Gold now has multiple closes below its 200-day moving average for the first time since early June 🚨 $GLD went into a technical correction the last time this happened 📉 📉 pic.twitter.com/DbGVgNNqTq

— Barchart (@Barchart) August 31, 2026
Renewed bets on a Federal Reserve rate hike have driven the latest leg down. Higher rates weigh on gold because the metal pays no yield.

Goldman Sachs and Fidelity Still Point HigherNonetheless, Goldman Sachs Research reaffirmed its 4,900 target for the end of 2026 in a note published August 28. From the press-time price, that implies roughly 10% upside.

The bank had cut that target by $500 in June as bets on 2026 rate cuts faded. The reduced figure still indicated gains, just smaller ones.

Senior commodities analyst Lina Thomas and Global Commodities Research co-head Daan Struyven anchored the call on official buying.

“We continue to see elevated central bank gold accumulation as a multi-year trend, as central banks diversify their reserves to hedge geopolitical and financial risks, consistent with recent survey evidence,” they wrote.

The bank expects central banks to buy an average of 50 tonnes per month in 2026, up from 17 tonnes before 2022. 

Meanwhile, Fidelity’s analysis valued gold around 5,000 against the global M2 money supply, about 13% above the press-time price. 

As for gold, it gained ground last week as the global liquidity profile has started to recover. Based on my Gold & Liquidity regression between global M2 and gold, gold is worth around $5k.

Below is a longer chart, which illustrates how gold has gone from a pure play on real… pic.twitter.com/oRkpLXZiI9

— Jurrien Timmer (@TimmerFidelity) August 14, 2026
The near-term risk is one Goldman already named. It’s June note put gold at 4,400 by year-end if the Fed hikes, and the metal traded there on Monday. A sustained break would also test the debasement trade, which ties gold and Bitcoin (BTC) demand to currency erosion

Subscribe to our YouTube channel to watch leaders and journalists provide expert insights

Investor sentiment in the cryptocurrency markets has recovered significantly in the last two weeks. The Crypto Fear and Greed Index, which was at 27 on August 12th and in the “fear” zone, rose to 74 on Tuesday, moving into the “greed” zone. The index then fell back to 65 on Wednesday.

The index reaching 74 marks its highest level since the period before the major market crash in October 2025. The indicator was last at similar levels on October 5, 2025. Just five days after that date, a sharp sell-off liquidated approximately $19 billion in leveraged positions.

The shift in market sentiment has also been reflected in the price performance of crypto assets. Bitcoin, which traded below $68,000 last week, quickly approached $80,000. Some major cryptocurrencies even saw increases of up to 70%.

Meme tokens have particularly stood out in the altcoin market. While Dogecoin gained approximately 24% in value over the past week, some smaller-scale meme coins with lower trading volumes saw even sharper increases. Thinking Cat rose 131%, Cash Cat 113%, while Dogecoin nearly doubled in value.

Analysts view the shift of capital towards small-cap tokens with low trading volume as a sign of a strong return in risk appetite. However, this same movement is also interpreted as increasing the risk of a correction due to market overheating.

The next major test for the markets is expected to be Federal Reserve Chairman Kevin Warsh’s first keynote address at the Jackson Hole meeting on Friday. Warsh’s messages regarding monetary policy are said to directly impact risk appetite in the cryptocurrency market.

*This is not investment advice.

Follow our Telegram and Twitter account now for exclusive news, analytics and on-chain data!

In brief ECB executive board member Piero Cipollone said the digital euro guarantees the maximum privacy current technology can offer, in an interview published Monday. Offline payments would be visible only to the payer and payee, while online the Eurosystem could not identify either party, though the banks involved could. The European Parliament agreed its negotiating position in July, with a pilot due in the second half of 2027 and first issuance targeted for 2029. A digital euro would give users as much privacy as current technology allows, a European Central Bank executive board member has said, in the bank's sharpest answer yet to the charge that a central bank digital currency would let Frankfurt watch how Europeans spend.

Piero Cipollone made the claim in an interview with Italian outlet ilsussidiario.net, conducted on August 10 and published by the ECB on Monday. Asked whether the bank could monitor payment habits and track how each citizen uses the currency, he said offline payments would run directly between individuals, with the details available only to the payer and the payee.

Online, he said, the Eurosystem would not be able to identify the people making or receiving payments. Only the banks involved in a transaction could do that, including for anti-money laundering purposes.

"The digital euro guarantees the maximum level of privacy that current technology can offer," Cipollone said.

Payments would still be visible to the commercial banks distributing the currency, which would keep the identity and reporting obligations they carry today. Privacy features are built in but do not exempt the currency from the rules applying to all money, an ECB spokesperson told Decrypt in December.

The European Parliament agreed its negotiating position on the regulation in July, and talks with member states are targeting a deal by the end of 2026. The ECB has named 36 payment providers, among them Deutsche Bank, UniCredit and Revolut, for a 12-month pilot from the second half of 2027, with first issuance targeted for 2029.

The EU, the US and CBDCsWashington reached the opposite conclusion on the same privacy question, and in July wrote it into law. The 21st Century ROAD to Housing Act bars the Federal Reserve from issuing a central bank digital currency until the end of 2030, after which it would need explicit authorization from Congress. The Senate passed the bill 85-5 in June, and it became law on July 11 without President Donald Trump's signature.

The ban exempts dollar-denominated currency that is open, permissionless and private, leaving the stablecoin issuers regulated under last year's GENIUS Act untouched.

Cipollone has argued that route carries a cost for Europe. In July he warned that growing stablecoin use would strip European banks of retail deposits, on top of the fees and transaction data they are already losing to mobile payment platforms.

Daily Debrief NewsletterStart every day with the top news stories right now, plus original features, a podcast, videos and more.

XRP’s 70% Rally Is Fading: Here’s the Level That Could Decide What Happens Next

GCSA Agent demonstrates autonomous vulnerability analysis and PoC generation capabilities on a highly challenging real-world vulnerability benchmark

The Global Cybersecurity Alliance (GCSA) today announced that GCSA Agent achieved a 91.3% success rate on the CyberGym benchmark, placing it within CyberGym’s “Leading Systems Above 90%” category.

CyberGym is a large-scale, real-world cybersecurity evaluation framework developed by a research team at the University of California, Berkeley. It contains 1,507 historical real-world vulnerability test cases across 188 major software projects and is designed to evaluate the practical capabilities of AI agents in real-world vulnerability analysis scenarios.

Unlike traditional AI benchmarks that primarily assess code understanding, knowledge-based question answering, or static analysis, CyberGym requires AI agents to work directly within real-world vulnerable code environments.

In its core Level 1 evaluation, an AI agent is provided only with a vulnerability description and an unpatched code repository. It must then autonomously perform code analysis, locate the vulnerability, reason about potential attack paths, construct a PoC, and execute it for validation. A task is considered successful only if the generated PoC successfully triggers the target vulnerability in the vulnerable version while failing to reproduce the issue in the patched version.

CyberGym therefore measures more than whether an AI system can simply “understand code.” It evaluates whether the AI can complete the full process from security analysis to vulnerability reproduction and validation.

From Large Language Models to Security AgentsIn this CyberGym evaluation, GCSA Agent operated on Grok 4.5 and Grok 4.6 models and achieved a final success rate of 91.3%.

The result also reflects an important shift taking place in AI cybersecurity:

The underlying large language model alone no longer determines the system’s ultimate security capabilities.

Real-world vulnerability research typically requires a continuous sequence of tasks, including understanding vulnerability descriptions, searching large codebases, identifying attack surfaces, formulating vulnerability hypotheses, generating test inputs, executing programs, analyzing feedback, and repeatedly iterating on PoCs.

GCSA Agent is built around an agentic security workflow designed to support this end-to-end process.

Its objective is not simply to use a large language model for code analysis, but to enable AI to operate within real execution environments, autonomously formulate hypotheses around security issues, collect runtime evidence, execute tests, and ultimately validate security findings through reproducible results.

The CyberGym evaluation provides a quantitative external benchmark for these capabilities.

Vulnerability Research Capabilities for the Real WorldA core value of CyberGym lies in narrowing the gap between traditional AI testing and real-world cybersecurity research.

Its evaluation environment restores software projects to their pre-patch vulnerable states. An AI agent may need to autonomously identify an issue within a large codebase containing thousands of files and millions of lines of code, and ultimately generate a PoC capable of actually triggering the vulnerability.

More importantly, further CyberGym research has shown that such agentic security capabilities are not limited to reproducing known vulnerabilities.

In open-ended vulnerability research experiments, AI agents have identified multiple previously unknown zero-day vulnerabilities as well as historical security patches that did not fully resolve the underlying vulnerabilities. These findings demonstrate the potential for autonomous vulnerability analysis technologies to evolve from reproducing known vulnerabilities toward discovering real-world security flaws.

For GCSA, this represents an even more important direction of development.

Benchmark performance is not the end goal.

GCSA aims to further develop AI Security Agents capable of operating in real-world cybersecurity environments and gradually participating across the full security lifecycle, from vulnerability discovery and analysis to validation and subsequent remediation.

Building AI-Native Cybersecurity CapabilitiesAs artificial intelligence accelerates software development, AI is also transforming the way vulnerabilities are researched and cyber threats are addressed.

As software systems continue to grow in scale and complexity, the next generation of cybersecurity will increasingly depend on collaboration between human security experts and autonomous AI agents.

AI Security Agents have the potential to help security teams:

Identify software vulnerabilities with genuine exploitation potential at an earlier stage;
Automatically analyse complex attack paths across large codebases;
Automatically generate PoCs and perform execution-level vulnerability validation;
Reduce false positives in traditional security detection through real execution results;
Accelerate vulnerability assessment, validation, and remediation;
Expand the scale of software and systems that specialised security teams are able to cover.
GCSA Agent’s 91.3% score on CyberGym represents an important milestone in GCSA’s development of AI-native cybersecurity capabilities.

Going forward, GCSA will continue advancing research into autonomous vulnerability analysis, AI Security Agents, and intelligent cybersecurity technologies, further translating frontier AI capabilities into real-world security capabilities and providing technical support for a safer, more trustworthy, and more resilient digital environment.

Source: GCSA Global Cybersecurity Alliance
Official Website: www.gcsa.org

Entry-level workers are taking the sharpest hit from artificial intelligence (AI), according to Goldman Sachs. The bank tracked hiring across developed economies and found the drag concentrated at the start of careers.

The bank published the research on Wednesday after examining employment growth across more than 800 occupations. 

Where AI Job Losses Show Up FirstIndustries most exposed to automation have posted slower growth in job openings since the second half of 2022. Goldman said the pattern is clearest in Germany, Australia, and the US.

Employment across information and communication services has cooled in almost every major developed economy over the same period. Outside the US, headcount in those industries still sits near or above its long-run trend.

Narrower categories tell a sharper story. Employment in call centers, software publishing, management consulting, and advertising services has fallen below the historical trend across developed markets.

According to the research, call centers stand out the most. Employment trails trend by 39% in the US, 33% in Canada, and 27% in Germany.

Goldman reads that as proof that the pressure lands first where automation tools already exist. Uber made that link explicit in July when it tied customer service job cuts to an AI efficiency push.

Adoption explains part of the gap. Major developed markets have reached rates of 15% to 20%, with France, the US, the Netherlands, and the UK leading. Italy, Japan, and New Zealand sit at the bottom, while emerging markets range between 10% and 15%.

Follow us on X to get the latest news as it happens

Entry-Level Workers Carry the CostThe damage concentrates at the start of careers. Across the wider workforce, the effect stays small. A 10% AI exposure was linked to a 0.1% drag on annual headcount growth in France, Canada, and the US.

For entry-level roles, that same exposure cut growth by more than 0.6 points in Australia. The US drag topped 0.2 points.

The signal is already reaching campuses, with students avoiding computer science degrees in favor of fields they judge safer. Earlier, research from Goldman Sachs estimated that the technology was subtracting 16,000 jobs per month from US payroll growth.

US Layoff Data Backs the TrendMeanwhile, Challenger, Gray & Christmas counted 33,429 job cuts in July. This was the lowest monthly total in two years.

AI led all stated reasons for the fifth consecutive month, accounting for 10,970 of those cuts, or 33%. Employers have named it in 112,713 announcements this year, roughly 24% of the total.

Technology sits at the center of the cutting with 149,023 announcements through July, up 67% from a year earlier. The sector now accounts for 31% of all 2026 cuts, extending the losses in tech and finance tracked earlier.

Andy Challenger, the firm’s chief revenue officer, said corporate messaging around the technology has shifted.

“Naming AI in a layoff announcement can win over investors while pushing current and prospective employees away. That’s why the messaging has swung from hedging to aggressively citing it,” he said.

Hiring plans complicate the picture. Companies announced 107,500 planned hires through July, up 25% from the same stretch of 2025. Demand is strongest in aerospace, energy, and manufacturing.

“Hiring has also increased over last year by 25%, so while AI is shifting the labor market, it is not dismantling it,” Challenger added.

Goldman concluded that the hiring pressure is visible worldwide but still confined to a narrow set of industries and workers. That containment is what US lawmakers demanding action on displacement will watch as adoption rises.

Subscribe to our YouTube channel to watch leaders and journalists provide expert insights

Pudgy Penguins (PENGU) Pushes Off the Lows: A Crucial Level to Hold First

Not financial or tax advice. PANews content is strictly educational and informational and is not investment advice, financial advice, tax advice, legal advice, or a solicitation to buy or sell any digital asset, security, or financial product. Do your own research and consult qualified advisers.

Disclosure. PANews may publish sponsored content, partner content, advertisements, affiliate links, event promotions, and market commentary involving Web3 projects, service providers, or financial products. PANews personnel, contributors, or affiliates may hold digital assets or other interests related to covered topics. See our Terms of Service.

This site is protected by reCAPTCHA.

Bitcoin Has Never Faced Global Bond Yields This High Since It Was Born

XRP Broke a Level It Hadn’t Touched in Over a Year — And Whales Bought It

Not financial or tax advice. PANews content is strictly educational and informational and is not investment advice, financial advice, tax advice, legal advice, or a solicitation to buy or sell any digital asset, security, or financial product. Do your own research and consult qualified advisers.

Disclosure. PANews may publish sponsored content, partner content, advertisements, affiliate links, event promotions, and market commentary involving Web3 projects, service providers, or financial products. PANews personnel, contributors, or affiliates may hold digital assets or other interests related to covered topics. See our Terms of Service.

This site is protected by reCAPTCHA.

Özellikle ABD’den gelen zayıf istihdam verileri, piyasada faiz artışı beklentilerinin azalmasına neden olurken bu durum Altın için yeniden pozitif bir ortam oluşturdu. Altındaki yükselişin en önemli nedenlerinden biri ise sadece bireysel yatırımcıların değil, merkez bankalarının da Altın almaya devam etmesi. Merkez bankaları son yıllarda rezervlerinde Dolar yerine Altın’ın ağırlığını artırmaya çalışıyor. Bunun temel sebebi ise Altın’ın herhangi bir ülkenin borcu olmaması ve ekonomik veya siyasi belirsizlik dönemlerinde güvenli liman olarak görülmesi. 2026 yılında yapılan bir ankette merkez bankalarının çok büyük bölümü Altın rezervlerinin önümüzdeki dönemde artmasını beklediğini belirtiyor.

İlginizi Çekebilir: Hyperliquid’de Toparlanma Sinyali: HYPE Kritik Dirençte!

Diğer taraftan dünya genelindeki siyasi ve ekonomik belirsizlikler de Altın’a olan ilgiyi canlı tutuyor. Özellikle ABD ekonomisinin ne kadar güçlü kalacağı, faizlerin hangi seviyede tutulacağı ve Orta Doğu’daki gelişmeler yatırımcıların güvenli liman arayışını artırıyor. Son dönemde ABD’den gelen zayıf iş gücü verileri de piyasada faiz beklentilerinin değişmesine neden olarak Altın’ın yeniden güç kazanmasını sağladı.

Teknik olarak incelersek:

XAU/USD paritesi günlük grafiği. 4000$ seviyelerinde yatay seyrettikten sonra tepki alıp yükseliş gösteren Altın, ilk olarak 4220$ bölgesini geçti ve üzerinde kapanış yaparak iç yapısındaki düşüşü sonlandırdı. Daha sonrasında bu alanı destek edinen Altın, önemli bir bölgeye gelmiş bulunuyor ve bu alanı da şuan kazanmış durumda. 4385$ üzerinde günlük kapanışların gelmesi Altın tarafında yükselişin devam edeceğinin işareti olacaktır. Bu bölgeyi de destek edinmesini bekleriz ve Altın için 4580$ seviyelerine kadar bir yükseliş görebiliriz. Altın tarafında düşüş sonrasında bir soğuma gerçekleştiğini ve bu alanların tekrar kazanılabileceğini bizlere göstermeye başladı.

4385$ üzerinde kalıcılık sağlanması halinde 4580$ seviyelerine kadar yükselmesini ve ilk direnç bölgesinin burası olmasını bekleyebiliriz. ABD tarafından gelen veriler şuanlık Altın’ın olumlu etkilenmesini sağlıyor. Özellikle piyasaların ABD enflasyon verisini beklediği bu dönemde, gelecek verilerin beklentileri desteklemesi halinde Altın tarafında yükselişin devam etmesini bekleyebiliriz. Şuan için 4385$ bölgesinin korunması, yükseliş senaryosunun devamı açısından önemli olacaktır.

Son dakika kripto para haberleri için hemen tıkla

Konu ile ilgili yorumlarınızı bize yazabilirsiniz. Ayrıca, bu tarz bilgilendirici içeriklerin devamının gelmesini isterseniz, bizleri Telegram, Youtube ve Twitter kanallarımızdan takip edebilirsiniz.

Not financial or tax advice. PANews content is strictly educational and informational and is not investment advice, financial advice, tax advice, legal advice, or a solicitation to buy or sell any digital asset, security, or financial product. Do your own research and consult qualified advisers.

Disclosure. PANews may publish sponsored content, partner content, advertisements, affiliate links, event promotions, and market commentary involving Web3 projects, service providers, or financial products. PANews personnel, contributors, or affiliates may hold digital assets or other interests related to covered topics. See our Terms of Service.

This site is protected by reCAPTCHA.

Analysis

Nearly 1 in 4 CHROs report entry-level hiring freezes tied to AI automation. But only 20% of organizations have seen significant value from the tools driving those cuts.

There is a quiet contradiction running through enterprise hiring right now. AI is everywhere in the corporate stack — 95% of organizations have implemented it in some form over the past year — yet only 1 in 5 say they have seen significant or transformational value from it. That gap between deployment and payoff has not stopped leaders from making permanent changes to the way they staff the bottom of the org chart.

According to a Gartner survey of 110 CHROs released July 27, 2026, 22% report that at least one business leader in their organization has stopped hiring for entry-level roles because of AI automation. Companies are dismantling their junior talent pipelines before they have validated that AI can actually replace the work those juniors were doing.

The Pattern Is Bigger Than One Survey This is not an isolated HR policy trend. It shows up in the macro labor data and in the age-specific employment numbers, and the picture it draws is consistent: AI is shifting the labor market, but the burden is falling unevenly.

The Challenger, Gray & Christmas July report, released August 6, puts AI as the number one reason for job cuts for the fifth consecutive month. July saw 33,429 announced cuts — the lowest monthly total in two years and down 46% year-over-year — with 10,970, or 33%, attributed to AI. Yet the same data shows announced hiring plans up 25% over last year, with 16,095 new positions announced in July alone. “Hiring has also increased over last year by 25%, so while AI is shifting the labor market, it is not dismantling it,” said Andy Challenger, the firm’s chief revenue officer.

So who is losing out? A Stanford SIEPR policy brief from July 2026 found that employment for 22-to-25-year-olds in AI-exposed occupations has declined since ChatGPT’s late-2022 launch, while employment for older, more experienced workers has remained stable or grown. Over 80% of employees report using AI, but only about 5% of firms report a measurable impact on employment levels. The junior tier is being hollowed out, even as firms struggle to prove that AI can do the work those juniors were hired to perform.

The Cost Paradox This creates a specific financial risk. Companies that freeze entry-level hiring today are betting they will not need to grow their own talent tomorrow. But that bet has a known failure mode.

Kaelyn Lowmaster, a director analyst in Gartner’s HR practice, warns that companies eliminating early-career pipelines tend to pay a premium later for experienced talent hired externally rather than developed from within. “Instead of eliminating these early career roles, organizations should redefine them to enable earlier contributions to higher-value work and build the talent they will need in the future,” she said.

The alternative — redefining rather than eliminating early-career roles — requires rethinking how junior employees actually contribute. Meaghan Kelly, also a director analyst in Gartner’s HR practice, frames it bluntly: “Traditional development approaches are no longer sufficient in an AI-enabled environment. Organizations can’t rely on gradual skill-building through routine work. Instead, they must provide the support structures that allow early career employees to operate effectively in more complex, judgment-intensive roles much earlier in their careers.”

Some companies are already testing this approach. Amazon plans to hire 11,000 interns and new college grads this year even as AWS builds and sells AI agents designed to automate recruiting, coding, and claims processing. AWS CEO Matt Garman argued on the Platformer podcast that mass job elimination would undercut the economy AI depends on: “The math doesn’t work out.” Amazon has cut roughly 30,000 corporate jobs since October, but Garman says those cuts came from flattening management layers, not from AI replacing the work itself. Amazon has more software developers today than it did two years ago despite heavy internal use of AI coding tools.

We have been tracking the enterprise agent deployment curve closely — from Cisco rolling out personalized AI agents to all 90,000 employees to the broader pattern of companies restructuring around agentic workflows. The efficiency case is real. But the Gartner data points to a gap between what companies are deploying and what they are proving. When nearly 1 in 4 CHROs report that their leadership has already frozen entry-level roles based on AI, and only 20% of organizations have seen significant value from those same AI investments, the pipeline destruction is running ahead of the evidence.

The industry is optimizing for efficiency before it has validated the replacement. If the math does not work out — if AI proves less capable of filling those junior roles than leaders assumed — the companies that cut their pipelines first will be the ones paying the most to rebuild them.

Ethoswarm Dana Ellison works for Forkast.
Minds can also work for you.

Minds are persistent AI beings with instincts, identity, and a job.
Awaken one on Ethoswarm.

Awaken your mind →

TLDR: Upbit BONK delisting will end BONK/KRW and BONK/USDT trading on September 7 after unresolved security and disclosure concerns. BONK withdrawals will stay available until October 7, giving Upbit users 30 days after trading ends to transfer tokens elsewhere. BONK fell 4.8% after the announcement and touched $0.00000255, marking its lowest reported price since November 2023 amid fresh selling. BONK has dropped 30.5% over one month and trades about 95.4% below its November 2024 all-time high of $0.00005825 after the latest decline. South Korean exchange Upbit will remove Bonk from its spot markets on September 7 after a month-long review. The Upbit BONK delisting affects the BONK/KRW and BONK/USDT pairs, ending access to a major Korean retail market. Upbit cited unresolved security concerns and inadequate disclosure of material information by the project. 

BONK reacted quickly, falling 4.8% to about $0.000002667 after the announcement. The token later touched $0.00000255, its lowest reported level since November 2023. BONK has also fallen 30.5% over one month, underperforming several larger meme coins during the same period.

Upbit BONK Delisting Follows Security and Disclosure Review Upbit said trading support will stop at 15:00 KST on September 7. The exchange will cancel every open order across the affected markets when trading ends. The decision follows BONK’s cautionary designation on July 7 and a review that lasted about one month.

Upbit to Delist BONK on September 7

South Korea’s largest crypto exchange, Upbit, will delist Bonk (BONK) on September 7, affecting the BONK/KRW and BONK/USDT pairs. Upbit cited unresolved security incidents and insufficient disclosure of material information. BONK withdrawals… pic.twitter.com/Z2YugwzkmC

— Wu Blockchain (@WuBlockchain) August 7, 2026

The Upbit BONK delisting centers on concerns that the exchange says were not resolved during that review. Upbit cited a security incident whose cause remained unidentified or inadequately addressed. It also pointed to delays in disclosing information that could materially affect traders and token holders.

The review followed reports of a $20 million governance attack involving the BONK DAO treasury. Upbit said the issues behind its cautionary designation were not sufficiently corrected. The exchange therefore moved to end trading under its asset review and trading support policies.

The move also removes BONK from one of South Korea’s largest retail trading venues. That could reduce local liquidity after September 7, especially for traders who rely on won-denominated markets. However, the decision does not immediately block holders from moving their tokens elsewhere.

BONK price weakness had already developed before the final notice. The meme coin has lost 30.5% over the past month. Dogecoin and Shiba Inu recorded smaller single-digit monthly declines over the same period. That performance left BONK among the weaker large meme coins during the recent broader market rebound.

Upbit BONK Delisting Leaves Holders Until October 7 to Exit BONK withdrawals will remain available through October 7, giving users 30 days after trading support ends. Holders can transfer tokens to another supported exchange or an external wallet before that deadline. Upbit warned that deposits made after the delisting will not be credited to user accounts.

The Upbit BONK delisting also changes how the crypto exchange will handle future token-related services. Upbit said it will not support airdrops, wallet upgrades, or hard forks for BONK after trading support ends. Users who send tokens incorrectly after the withdrawal window could also face lengthy recovery procedures.

BONK price action shows how sharply the token has retreated from its 2024 peak. BONK reached an all-time high near $0.00005825 in November 2024. At about $0.000002667, the token trades roughly 95.4% below that level.

BONK price The Upbit BONK delisting may also shift attention toward other Korean exchanges that still support the token. Any similar listing reviews could affect domestic access and liquidity. For now, Upbit users have until September 7 to trade BONK and until October 7 to withdraw it.

The timing gives holders a defined transition period, but it also separates trading from withdrawals. Selling on Upbit ends first, while transfers remain available for another month. That distinction matters for users who intend to retain BONK rather than convert it before the local market closes.

The final weeks before the trading cutoff may bring higher turnover as holders adjust positions. Open orders will disappear when trading support ends. Traders who plan to keep BONK must therefore arrange withdrawals before Upbit closes the remaining transfer window.

Not financial or tax advice. PANews content is strictly educational and informational and is not investment advice, financial advice, tax advice, legal advice, or a solicitation to buy or sell any digital asset, security, or financial product. Do your own research and consult qualified advisers.

Disclosure. PANews may publish sponsored content, partner content, advertisements, affiliate links, event promotions, and market commentary involving Web3 projects, service providers, or financial products. PANews personnel, contributors, or affiliates may hold digital assets or other interests related to covered topics. See our Terms of Service.

This site is protected by reCAPTCHA.

Trading activity in the cryptocurrency market has slowed significantly, with weekly trading volume falling to its lowest level since 2026. According to the latest data shared by Kaiko, a data and analysis provider for digital asset markets, the total trading volume in the cryptocurrency market last week was approximately $15 billion. This figure represents a significant decline compared to levels at the beginning of the year.

According to Kaiko’s analysis, the highest trading volume of the year was recorded in January. At that time, the total volume in the markets was approximately three times the current level. According to the latest data, trading volume has decreased by approximately 70% compared to the peak level in January.

Experts say the decline in trading volume reflects a weakening of investors’ risk appetite. They note that in a low-volume environment, price movements in the market can become more fragile, and even relatively small buy or sell orders can increase volatility. Therefore, investors are advised to be more cautious about sudden price fluctuations during periods of low liquidity.

According to market analysts, there are multiple reasons behind the decline in trading volume. Ongoing uncertainties in the global economy, cautious expectations regarding central bank monetary policies, increased selectivity among institutional investors, and the seasonal slowdown in trading traditionally seen during the summer months are among the main factors supporting this decline.

However, experts also point out that low trading volume alone is not an indicator determining the long-term direction of the market. They note that historically, periods of prolonged low volume in the cryptocurrency market have been followed by a return to strong trading activity.

Funding flows, particularly into spot Bitcoin and Ethereum ETFs, along with macroeconomic developments and upcoming regulatory steps, are expected to play a decisive role in the renewed increase in trading volumes in the coming period.

Analysts say that trading volumes could recover if investor confidence strengthens and new capital inflows into the market accelerate. However, the current outlook indicates that investors in the cryptocurrency market remain cautious, and trading activity has fallen to one of its weakest periods of the year.

*This is not investment advice.

Follow our Telegram and Twitter account now for exclusive news, analytics and on-chain data!