Centralized exchange Gemini recently announced that it parted ways with three senior executives. The leadership changes come amid broader operational cutbacks and workforce reductions.

Following the announcement, the company’s shares declined further, extending a downward trend that has persisted since Gemini went public last September. The latest developments have prompted renewed scrutiny over the exchange’s long-term outlook.

Sponsored

Executive Shakeup Follows Deep Cuts

In a recent blog post, Tyler and Cameron Winklevoss announced that Gemini had parted ways with its Chief Financial Officer, Chief Legal Officer, and Chief Operating Officer. They said interim replacements had been appointed for the CFO and CLO roles, while the COO position would not be filled.

The founders characterized the changes as part of a broader transformation at the company, referring to the initiative as “Gemini 2.0.” They noted that recent developments in the crypto industry have influenced this shift.

“During this time, but really more recently, rapid breakthroughs in AI have begun to dramatically transform the way we work at Gemini. Simultaneously, the advent of prediction markets has begun to dramatically transform marketplaces, including our own,” the blog post stated.

The announcement drew heightened attention as it followed Gemini’s decision weeks earlier to reduce its global workforce by 25%. In addition, Gemini has exited several international markets, including the United Kingdom, the European Union, and Australia.

The latest developments prompted renewed volatility in the company’s stock, extending a steep decline that has persisted since its September listing. Investors who purchased GEMI at its $28 IPO price are now facing losses of roughly 77%.

Sponsored

In a recent SEC filing, the company also disclosed an estimated net loss of approximately $595 million for 2025.

Taken together, these developments have intensified scrutiny of the exchange’s valuation.

Public Markets Reprice Gemini Growth

The sharp repricing of Gemini’s stock has renewed debate over whether the exchange was fundamentally overvalued at its initial public offering (IPO).

Sponsored

Its initial valuation reflected expectations of sustained trading volumes and revenue expansion. Given the cyclical nature of the crypto market, pricing may have been influenced by elevated trading activity and heightened retail participation. 

The subsequent decline, unfolding amid a broader market downturn, suggests a reassessment of earnings expectations. 

The developments also highlight intensifying competitive pressures between centralized exchanges. 

Market share and liquidity remain concentrated among larger platforms with deeper order books and stronger network effects. Meanwhile, mid-tier exchanges face rising fixed costs without equivalent trading scale to support margins.

Recent data from CoinGecko supports the situation. 

Sponsored

In a January report on centralized exchange market share by trading volume, the data aggregator found that in 2025, Binance accounted for 39.2% of total spot volume among the top exchanges, processing $7.3 trillion in volume. Other leading platforms, including Bybit, MEXC, and Coinbase, also maintained meaningful shares of global volume.

Gemini did not place among the top 10. According to CoinMarketCap data, the exchange currently ranks 24th, with a 24-hour trading volume of $54 million. 

Within that context, workforce reductions and geographic pullbacks may represent cost-control measures and strategic adjustments to an increasingly consolidated market. 

How Gemini executes this transition will likely shape whether shareholders view the current turbulence as a short-term adjustment or a sign of longer-term structural challenges.

OpenAI has introduced a new system called EVMbench, designed to measure how well artificial intelligence agents can find and fix security flaws in crypto smart contracts.

The company announced on Feb. 18 that it has developed EVMbench in partnership with Paradigm. The benchmark focuses on contracts built for the Ethereum Virtual Machine and is meant to test how AI systems perform in real financial settings.

OpenAI said smart contracts currently secure more than $100 billion in open-source crypto assets, making security testing increasingly important as AI tools become more capable.

Testing how AI handles real security risks

EVMbench evaluates AI agents across three main tasks: detecting vulnerabilities, fixing flawed code, and carrying out simulated attacks. The system is built using 120 high-risk issues drawn from 40 past security audits, many of them from public auditing competitions.

Additional scenarios were taken from reviews of the Tempo blockchain, a payments-focused network designed for stablecoin use. These cases were added to reflect how smart contracts are used in financial applications.

To build the test environment, OpenAI adapted existing exploit scripts and created new ones where needed. All exploit tests run in isolated systems rather than on live networks, and only previously disclosed vulnerabilities are included.

In detection mode, agents review contract code and try to identify known security flaws. In patch mode, they must fix those flaws without breaking the software. In exploit mode, agents attempt to drain funds from vulnerable contracts in a controlled setting.

Early results and industry impact

OpenAI said a custom testing framework was developed to ensure results can be reproduced and verified.

The company tested several advanced models using EVMbench. In exploit mode, GPT-5.3-Codex achieved a score of 72.2%, compared with 31.9% for GPT-5, released six months earlier. Detection and patching scores were lower, showing that many vulnerabilities are still difficult for AI systems to handle.

Researchers observed that agents performed best when goals were clear, such as draining funds. Performance dropped when tasks required deeper analysis, such as reviewing large codebases or fixing subtle bugs.

OpenAI acknowledged that EVMbench does not fully reflect real-world conditions. Many major crypto projects undergo more extensive reviews than those included in the dataset. Some timing-based and multi-chain attacks are also outside the system’s scope.

The company said the benchmark is intended to support defensive use of AI in cybersecurity. As AI tools become more powerful, they could be used by both attackers and auditors. Measuring their capabilities is seen as a way to reduce risk and encourage responsible deployment.

Alongside the release, OpenAI said it is expanding security programs and investing $10 million in API credits to support open-source and infrastructure protection. All EVMbench tools and datasets have been made public to support further research.

Coinbase is expanding its crypto-backed lending product in the U.S. to include XRP, DOGE$0.1017, Cardano’s ADA and LTC$53.44, widening access to a service it has pitched as a way for customers to unlock liquidity without selling their holdings.

The product allows users to post crypto as collateral and borrow up to $100,000 in Circle’s USDC stablecoin. The loans are routed through Morpho, a decentralized lending protocol, meaning the borrowing mechanics are handled on-chain rather than through Coinbase’s own balance sheet.

The service is available across the U.S., excluding New York.

The move brings some of crypto’s most retail-heavy tokens into a product that previously focused on bitcoin and ether. While Ethereum and Cardano holders can already earn yield through staking on their native networks, assets like XRP, DOGE and Litecoin do not offer built-in reward mechanisms.

For those investors, borrowing against their holdings has become one of the few ways to access liquidity without exiting the position.

Coinbase is also expanding the potential pool of collateral on its platform. The exchange reported it held $17.2 billion worth of XRP as of Dec. 31, according to an SEC filing, making the token one of the larger assets sitting in customer accounts.

Crypto-backed loans have long been marketed as a tax-efficient strategy, since borrowing against an asset does not trigger capital gains in the same way selling does.

But the structure comes with sharp risks when markets move quickly. If the value of the collateral falls too far relative to the loan, the position can be liquidated, meaning a third party can repay the debt and seize the collateral at a discount.

Coinbase applies an extra buffer when users take out a loan to reduce liquidation risk and sends notifications as the threshold is approached. Still, the exchange has also warned that collateral used through the product is wrapped, a process that allows tokens like XRP to exist on Ethereum-compatible networks.

Bitcoin is finding space to bounce, but not yet the fuel to run.

The macro backdrop has improved just enough to give bulls something to work with. Cooling headline inflation has strengthened expectations for three rate cuts this year, reviving the familiar playbook in which easier monetary policy supports risk assets.

And it could signal the possibility of liquidity slowly returning after months of tight financial conditions for crypto markets.

But caution against reading too much into that shift. The Federal Reserve is unlikely to embark on an aggressive easing cycle. Instead, it appears set for a measured approach that rebuilds liquidity gradually. That creates an environment where bitcoin can stage tactical rallies yet struggle to hold them.

Bitfinex analysts describe the market as one prone to moves in waves rather than clean breakouts.

“In this environment, volatility remains likely,” the firm said in a note shared with CoinDesk. “Tactical upside moves can occur when positioning becomes overly defensive, but a durable structural advance will require clearer confirmation from both macro disinflation trends and sustained spot demand.”

Spot recoveries continue to meet steady selling. Each bounce is absorbed more smoothly than earlier in the quarter, suggesting some stabilization.

The overnight tape is a good example. Bitcoin traded as high as $68,500 before rolling over during the U.S. afternoon and sliding under $66,000, a move that lined up with a stronger dollar and hawkish Fed minutes. That kind of intraday reversal is the market’s way of saying rallies are still fragile, and that traders are quick to sell the moment macro conditions turn even slightly less friendly.

“It is alarming that Bitcoin’s dynamics mirror the recent strengthening of the dollar. When investors become convinced that the rise of the dollar is a trend, there may be a sharp increase in volatility,” Alex Kuptsikevich, the FxPro chief market analyst, said in an email.”

“Volatility seems to have been turned off in this market, while stock indices are much livelier. There, investors are actively buying up dips, relying on support in the form of important moving averages: 50-day for the Dow Jones and Russell 2000 and 200-day for the Nasdaq100. The crypto market is now below its 50- and 200-day curves by 17% and 31%, respectively,” he added.

Sentiment remains fragile, meanwhile, as a crypto fear gauge has printed single digits on nine of the past fourteen days, territory rarely seen outside prior cycle lows.

At the same time, stablecoin outflows from major exchanges point to tighter liquidity, and long-term holders have shown signs of stress comparable to late bear-market phases in 2022, according to Glassnode.

For now, bitcoin appears caught between improving macro optics and stubborn supply. Tactical upside remains possible, especially when positioning leans too defensive.

A durable advance, however, likely requires clearer evidence of disinflation, a softer dollar and consistent spot demand. Until then, the path higher may be uneven.

Editor’s note: Global Games Show Riyadh 2026 signals a turning point for Saudi Arabia’s digital entertainment ecosystem as the kingdom accelerates growth across gaming, esports, and Web3. This press release outlines a multi-day program that combines live demonstrations, developer workshops, and high-profile panels, underscoring Riyadh’s emergence as a regional hub for interactive technology. The show also reinforces collaboration among startups, creators, and investors through dedicated networking spaces and matchmaking sessions. By bringing together leaders from across the industry, the event aims to catalyze partnerships and accelerate the creative economy envisioned in Vision 2030.

Key points

• Global Games Show Riyadh 2026 brings together gaming, esports, and Web3 within Saudi Vision 2030.
• The event features live demos, workshops, panels, and networking with industry leaders, indie developers to global publishers.
• It is organized by VAP Group and powered by Times of Games, with parallel events Global AI Show and Global Blockchain Show on a single ticket.

Why this matters

By concentrating expertise and investment in Riyadh, the Global Games Show aims to accelerate Saudi Arabia’s creative economy and position the Kingdom as a regional and global hub for interactive technology. The conference highlights trends in immersive gaming, cloud gaming, and monetization strategies, and emphasizes collaboration across startups, developers, and publishers, aligning with Vision 2030’s diversification goals.

What to watch next

• Updates on Day 1 and Day 2 sessions and key speakers.
• Public announcements of participating companies and partnerships.
• Ticketing details for the Global AI Show and Global Blockchain Show, accessible with one ticket.

Disclosure: The content below is a press release provided by the company/PR representative. It is published for informational purposes.

Global Games Show Riyadh 2026 : Fueling Saudi Arabia’s Vision 2030

Global Games Show Riyadh 2026 Riyadh edition is poised to become the ultimate destination for gaming enthusiasts, developers, and investors alike. Organized by VAP Group and powered by the Times of Games, the event promises a vibrant lineup of discussions and engaging experiences that symbolize the rapidly changing gaming sphere.

Participants can explore the latest in game development, esports, and interactive entertainment, with live demonstrations, workshops, and panels led by industry leaders. From indie developers to global publishers, companies will present their most innovative games and technologies, providing attendees with insights into the future of gaming.

Educational and strategic sessions focus on trends such as immersive gaming, cloud gaming, and monetization strategies. These discussions equip participants with knowledge to navigate challenges, leverage opportunities, and scale their ventures effectively.

Day 1 is all about the future of gaming technology, with talk on Saudi Arabia becoming a world esports capital, the next phase of gaming engines with Unreal Engine 6, brain–computer interfaces, and AI-generated game design. Experts will also discuss what the future of esports will look like in the Kingdom and how it is increasingly driving Vision 2030’s creative economy.

Day 2, entitled “Gameconomics,” explores the gaming business—from crowdfunded games to mobile gaming opportunity, player-coined communities, and developer–investor partnerships that form industry expansion.

By bringing a diverse mix of professionals under one roof, the Global Games Show strengthens Riyadh’s position as a hub for interactive technology and digital entertainment. Attendees also get access to other parallel flagship events, the Global AI Show and the Global Blockchain Show with just one ticket. GGS is a convergence of ideas, creativity, and opportunity in the gaming world.

Media enquiries :

Press contact : media@globalblockchainshow.com

Moonwell’s lending pools racked up about $1.78M in bad debt after a cbETH oracle mispriced the token at nearly $1 instead of around $2.2k, enabling bots and liquidators to drain collateral within hours of a misconfigured Chainlink-based update reportedly using AI-generated logic.

Decentralized finance lending protocol Moonwell suffered a $1.78 million exploit due to a pricing oracle bug that misvalued Coinbase-wrapped ETH (cbETH), according to reports from the platform.

The vulnerability originated in oracle calculation logic reportedly generated by the AI model Claude Opus 4.6, which introduced an incorrect scaling factor in the asset price feed, according to the protocol’s disclosure. Attackers borrowed against severely underpriced collateral, extracting funds before the error was detected and corrected.

The cbETH mispricing effectively collapsed the collateral requirement for borrowing within affected pools. Because lending systems rely on accurate collateral ratios, the incorrect price allowed attackers to extract assets with minimal backing value, according to the protocol’s technical analysis.

Price oracles represent critical security components in DeFi lending systems. Incorrect asset valuation can enable under-collateralized borrowing or liquidation failures. Many major DeFi exploits have historically involved oracle manipulation or pricing errors rather than core protocol flaws, according to industry security reports.

The Moonwell incident differs from traditional oracle exploits in that the faulty logic appears linked to automated AI code generation rather than malicious oracle data feeds, according to the protocol’s preliminary investigation.

The exploit highlights risks associated with AI-assisted smart-contract development in financial applications. Language models can accelerate coding workflows, but financial protocols require precise numerical correctness, unit handling and edge-case validation, according to blockchain security experts.

In DeFi systems, small arithmetic or scaling mistakes can translate into systemic vulnerabilities affecting collateral valuation and solvency. The incident raises questions about whether AI-generated contract components may require stricter auditing standards than manually written code, according to security researchers.

AI-assisted development is increasingly used across Web3 engineering workflows, from contract templates to integration logic. Security models and audit frameworks have not yet fully adapted to AI-generated contract code, according to industry observers.

The broader implications center on how automated code generation errors in financial logic represent a new category of DeFi risk. Oracle math, scaling factors and unit conversions remain high-precision domains where automation failures can propagate into protocol-level vulnerabilities, according to technical analysis of the incident.

As AI-assisted smart-contract development expands, audit methodologies will likely need to evolve toward verifying not only code correctness but generation provenance and numerical invariants, according to blockchain security firms.

Three researchers at the US Federal Reserve argue that prediction market Kalshi can better measure macroeconomic expectations in real time than existing solutions and thus should be incorporated into the Fed’s decision-making process.

The “Kalshi and the Rise of Macro Markets” paper was released on Feb. 12 by Federal Reserve Board principal economist Anthony Diercks, Federal Reserve research assistant Jared Dean Katz and Johns Hopkins research associate Jonathan Wright.

Kalshi data was compared with traditional surveys and market-implied forecasts to examine how beliefs about future economic outcomes change in response to macroeconomic news and statements from policymakers.

“Managing expectations is central to modern macroeconomic policy. Yet the tools that are often relied upon—surveys and financial derivatives—have many drawbacks,” the researchers said, adding that Kalshi can capture the market’s “beliefs directly and in real time.”

“Kalshi markets provide a high-frequency, continuously updated, distributionally rich benchmark that is valuable to both researchers and policymakers.”

Kalshi traders can bet on a range of markets tied to the Federal Reserve’s decision-making, including consumer price index inflation and payroll, in addition to other macroeconomic outcomes such as gross domestic product growth and gas prices.

The Fed researchers said Kalshi data should be used to provide a risk-neutral probability density function, which shows all possible outcomes of Fed interest rate decisions and how likely each one is. 

“Overall, we argue that Kalshi should be used to provide risk-neutral [probability density functions] concerning FOMC decisions at specific meetings” arguing that the current benchmark is “too far removed from the monetary policy interest rate decision.”

However, Fed research papers are only “preliminary materials circulated to stimulate discussion” and do not impact the central bank’s decision-making.