DD
MM
YYYY

PAGES

DD
MM
YYYY

spot_img

PAGES

Home Blog Page 8

ByteDance Vows Stronger Safeguards on Seedance 2.0 After Disney, Paramount Skydance Cease-and-Desist Letters Over Alleged IP Infringement

0

ByteDance announced on Monday that it is taking steps to strengthen safeguards on its newly launched AI video generator Seedance 2.0 to prevent unauthorized use of intellectual property, following cease-and-desist letters from major U.S. studios, including Disney and Paramount Skydance.

The move comes just days after Seedance 2.0 went viral in China for generating highly realistic cinematic content—including scenes featuring Tom Cruise and Brad Pitt in a fight—drawing widespread comparisons to OpenAI’s Sora and DeepSeek’s video capabilities.

Disney’s letter, first reported by Axios on Sunday, accused ByteDance of using copyrighted characters from Star Wars, Marvel, and other franchises to train and power Seedance 2.0 without permission. A source familiar with the matter told Reuters that Disney alleged ByteDance had “pre-packaged” the model with a pirated library of characters, treating them as public-domain clip art. The letter claimed Seedance was reproducing, distributing, and creating derivative works featuring Spider-Man, Darth Vader, and other protected figures.

Paramount Skydance sent a similar cease-and-desist letter, accusing ByteDance of “blatant infringement” of its intellectual property, according to Variety over the weekend. ByteDance’s statement did not name specific studios or elaborate on the measures being implemented, saying only: “We are taking steps to strengthen current safeguards as we work to prevent the unauthorized use of intellectual property and likeness by users.”

The rapid response reflects the high legal and reputational stakes for ByteDance as it pushes Seedance 2.0—praised for controllability, speed, production efficiency, and polished first-try results—into both domestic and international markets. The model’s viral success during the early days of Lunar New Year has amplified attention, but also exposed it to immediate scrutiny from Hollywood studios protective of their IP.

Disney’s action follows a pattern. In recent months, the studio sent similar demands to Character.AI to stop unauthorized use of its characters. Notably, Disney signed a licensing deal with OpenAI in December 2025, granting the U.S. company rights to use Star Wars, Pixar, and Marvel characters in its Sora video generator—highlighting a stark contrast in how studios are willing to engage with U.S. versus Chinese AI firms.

Seedance 2.0’s ability to generate realistic videos from text prompts, images, or reference clips has drawn praise from creators. Stockholm-based creative advertising executive Billy Boman told CNBC the progress in AI video generation over the past two years has been “nothing short of exceptional.” Hugging Face researcher Adina Yakefu called Seedance 2.0 “one of the most well-rounded video generation models I’ve tested,” noting its ability to deliver satisfying results on simple prompts with polished visuals, music, and cinematography.

However, the controversy over IP infringement underscores a broader challenge for Chinese AI video generators. Hollywood studios view unauthorized use of their characters, likenesses, and copyrighted material as a direct threat to licensing revenue and brand control. ByteDance’s open-ended prompt flexibility—allowing users to generate content featuring well-known figures—has fueled viral clips but also invited swift legal pushback.

The timing is particularly sensitive as Seedance 2.0’s release and rapid viral spread occurred just before and during Lunar New Year, a high-engagement period when family sharing and social media usage spike. ByteDance appears to have aimed to dominate the conversation during the holiday, but the IP allegations risk overshadowing the technological achievement.

ByteDance has not yet responded publicly to the specific allegations from Disney and Paramount Skydance. The company’s statement focused on future safeguards rather than addressing past or current content generated by users.

The episode highlights growing friction between Chinese AI labs and Western content owners. While ByteDance, Alibaba (with RynnBrain), and Kuaishou (with Kling 3.0) have made rapid advances in video generation and embodied AI, U.S. studios are increasingly willing to enforce IP rights aggressively—especially when models can produce high-quality derivative works featuring protected characters. Disney’s licensing deal with OpenAI contrasts sharply with its stance toward Chinese firms, suggesting studios may be more open to partnerships with U.S. companies that can offer enforceable compliance and revenue-sharing arrangements.

The path forward will likely involve tighter content filters, stricter prompt moderation, and potentially licensing negotiations if ByteDance seeks broader international adoption.

The controversy arises as China’s AI video generation sector heats up. Seedance 2.0, Kling 3.0, and emerging competitors are closing the gap with OpenAI’s Sora in realism, controllability, and creative flexibility—often at lower cost and with fewer restrictions.

Hyperliquid Is A Powerful Signal of DeFi Maturity and Efficiency Advantages 

0

Hyperliquid; a decentralized perpetuals-focused exchange on its own L1 chain has shown impressive growth and has surpassed Coinbase in certain metrics—particularly notional trading volume—the company’s actual revenue figures tell a different story.

Hyperliquid (protocol fees/revenue, per DefiLlama and Artemis): Annualized around $999 million to $1.11 billion recently, with 2025 totals in the $800–$844 million range. Recent daily revenue has hit peaks like $6.84 million (early Feb 2026), with 24h figures around $2–$2.6 million.

Much of this funds HYPE token buybacks rather than traditional profit. Coinbase full-year 2025 reported: Total revenue $7.18 billion up 9% YoY, including transaction fees, subscriptions/services ($2.8–$2.83 billion), and other streams. This is significantly higher than Hyperliquid’s protocol revenue.

Coinbase’s Q4 2025 alone was ~$1.78 billion, though it included a GAAP net loss due to crypto investment marks and expenses.
Hyperliquid’s revenue comes almost entirely from trading fees on perps/spot with low fees but massive volume, while Coinbase diversifies across retail/institutional trading, staking, custody, USDC interest, subscriptions, and more.

Coinbase is a publicly traded company with broader operational scale, compliance costs, and profitability challenges in volatile markets, but its top-line revenue dwarfs Hyperliquid’s. Hyperliquid is “quietly outgrowing” Coinbase in derivatives-specific activity and efficiency for crypto-native traders (low fees, on-chain transparency, high leverage).

Its HYPE token has outperformed COIN stock in early 2026 performance; +31.7% vs. -27% YTD at times, reflecting market excitement about decentralized perps. However, “making more money” typically refers to revenue/profit for the entity/exchange, not just volume or token gains—and Coinbase clearly leads there.

Hyperliquid’s model (decentralized, token buybacks via fees) differs from Coinbase’s centralized, regulated business, so direct apples-to-apples comparisons have limits.

Hyperliquid “makes more money” than Coinbase primarily stems from its dominance in notional trading volume and protocol fees in the decentralized perpetuals space, but the broader impacts of this shift especially in early 2026 are significant for the crypto industry, traders, centralized exchanges (CEXs), and decentralized finance (DeFi).

Hyperliquid has not surpassed Coinbase in total revenue. Coinbase reported ~$7.18 billion for full-year 2025 with subscription/services at ~$2.83 billion and transaction fees making up much of the rest, while Hyperliquid’s protocol fees were around $822–$844 million in 2025, with YTD 2026 figures in the $79 million range and daily/annualized peaks occasionally hitting higher.

However, Hyperliquid’s efficiency (massive volume with a tiny team of ~11–15 people vs. Coinbase’s thousands) and revenue-per-employee metrics highlight a lean, high-margin model.

Traders increasingly prefer on-chain platforms for perpetual futures due to lower fees, non-custodial control (no account freezes or KYC hassles for many users), transparency, and high leverage up to 50x without intermediaries.

This signals a structural migration: Hyperliquid captures users frustrated with CEX restrictions like US perp bans, outages, or regulatory risks. It’s “quietly outgrowing” Coinbase in derivatives-specific activity, forcing a reevaluation of where serious trading happens.

Liquidity is becoming more distributed, with on-chain venues challenging CEX dominance in perps; Hyperliquid now handles significant shares vs. Bybit and OKX in some metrics. Pressure on centralized Elexchanges especially Coinbase. Coinbase faces competition in derivatives and spot-like activity, contributing to challenges like its Q4 2025 revenue miss, and COIN stock underperformance -27% YTD early 2026 vs. HYPE +31.7%.

CEXs may need to innovate or risk losing market share to efficient DEXs. Coinbase has listed HYPE and enabled related trading, showing adaptation rather than outright rivalry. Hyperliquid generates substantial fees with minimal staff, while Coinbase’s scale brings compliance costs and slower innovation.

Hyperliquid’s model funnels ~97% of fees into HYPE buybacks/burns, creating deflationary pressure and direct value accrual to holders. This contrasts with Coinbase’s traditional equity structure. HYPE outperforms COIN significantly in early 2026 performance, reflecting market excitement for decentralized, revenue-sharing protocols over regulated incumbents.

Boosts confidence in “real revenue” DeFi projects (trading fees, not emissions/taxes), with institutions eyeing HYPE via ETF filings despite no VC allocation. Hyperliquid dominates on-chain perps, expands into prediction markets/options via HIP-4, no-liquidation designs, and tests new products, pressuring competitors like Polymarket, dYdX, or Aster/Lighter.

CEXs and DEXs coexist/compete more intensely, driving better UX, lower costs, and composability. Growth draws scrutiny, but also highlights DeFi’s appeal in restricted regions/markets. Hyperliquid’s model has criticisms e.g., liquidity provider risks during liquidations, structural conflicts where fees benefit from volatility/losses.

Notional volume inflates due to leverage—actual economic activity differs from spot-focused Coinbase. Early 2026 bearish conditions hit COIN harder, but Hyperliquid’s resilience shows strength. Hyperliquid’s rise isn’t about fully eclipsing Coinbase’s revenue scale yet—it’s a powerful signal of DeFi maturity, efficiency advantages, and trader migration to decentralized models.

This pressures incumbents to evolve, accelerates on-chain adoption, and underscores crypto’s shift toward transparent, high-performance trading infrastructure. If trends continue, it could reshape exchange hierarchies far beyond just perps.

Goldman Sachs Highlights that Recent US Stock Market Selloff is Likely Not Yet Over 

0
The logo for Goldman Sachs is seen on the trading floor at the New York Stock Exchange (NYSE) in New York City, New York, U.S., November 17, 2021. REUTERS/Andrew Kelly/Files

Goldman Sachs recently issued a warning indicating that the recent US stock market selloff is likely not over yet, based on analysis from their trading desk.

The key points from their note primarily via Bloomberg; US stocks rebounded late last week (Friday), nearly recovering from a sharp mid-week decline, but conditions remain choppy. Trend-following algorithmic funds specifically Commodity Trading Advisers, or CTAs — systematic strategies that trade based on price momentum rather than fundamentals have already been triggered to sell due to the S&P 500 breaching short-term levels.

Goldman expects these funds to stay net sellers over the coming week, potentially regardless of short-term market direction. A renewed decline could prompt about $33 billion in additional selling of US equities this week.

If selling pressure persists and the S&P 500 falls below 6,707, it could unlock up to $80 billion in further systematic selling over the next month. Goldman’s proprietary Panic Index which tracks factors like one-month S&P implied volatility and VIX-related measures surged to 9.22 last week — approaching “max fear” levels — signaling elevated investor stress and thin liquidity, which could amplify volatility.

Traders advised investors to “buckle up” for potential continued turbulence. This warning appears tied to a volatile period in early February 2026, including a tech/AI-related selloff earlier in the month that hit hedge funds and crowded trades hard.

Broader 2026 outlooks from Goldman Sachs Research issued earlier in January had been more constructive overall — forecasting solid global growth ~2.8%, US outperformance, and ~11% potential returns for global equities over the next 12 months — but this trading desk update highlights near-term tactical risks from momentum-driven flows.

Markets can shift quickly, so this reflects sentiment and positioning as of early February 2026 rather than a definitive long-term call. Investors are watching key S&P levels closely for confirmation of whether the selling intensifies or stabilizes.

Goldman Sachs’ warning about the ongoing market selloff has particularly significant implications for tech stocks, which have been at the epicenter of the recent volatility in early February 2026. The selloff was primarily triggered by fears of AI disruption to traditional software and tech business models, sparked by Anthropic’s release of an advanced AI automation tool.

This led to a sharp rotation out of tech into more defensive sectors like consumer staples. Goldman Sachs highlighted that software stocks entered a bear market, with their proprietary basket losing around $2 trillion from 2025 highs roughly a 30% drop. Many names saw massive declines: Oracle and Salesforce down ~27%.

Others like ServiceNow, Workday, SAP, and newer IPOs; Figma down 41% YTD. The iShares Expanded Tech-Software Sector ETF (IGV) dropped over 12-13% in recent sessions and entered oversold territory. The Nasdaq Composite saw sharp declines early in the month down ~1.6% on some days, with multi-day routs, underperforming the S&P 500.

The S&P 500 tech sector (.SPLRCT) and software services index faced steep losses before partial rebounds. Magnificent 7 and AI-related names: While not all details specify uniform hits, the group lagged amid concerns over massive AI capex and disruption risks.

Hedge funds reduced exposure to megacaps, and crowded trades unwound. Goldman noted hedge funds focused on tech/telecom/media suffered their worst days in nearly a year down up to 2.78% in a session. Trend-following CTAs (already selling after S&P breaches) could add $33 billion in equity sales with up to $80 billion more over the next month if S&P 500 drops below ~6,707.

Tech’s high weighting in indices means it bears much of this pressure. Markets saw a sharp rebound with the S&P 500 up ~2%, Nasdaq gaining, and tech/software clawing back. This extended into early the following week for some gains, as the pullback was viewed as oversold and overdone by some.

However, Goldman’s trading desk emphasized the selloff isn’t over yet — CTAs remain net sellers near-term (regardless of direction), liquidity is thin, and their Panic Index hit near “max fear” levels (9.22).

Volatility could persist or intensify if key levels break, with tech vulnerable due to its momentum-driven nature and ongoing AI uncertainty. While a tactical bounce is possible; some analysts see counter-trend rallies in oversold software, the near-term risk skews toward more turbulence for tech stocks if systematic flows accelerate.

Broader rotation to “real economy” sectors continues, but tech’s leadership could return if disruption fears ease or earnings prove resilient. Watch S&P levels closely — especially around 6,707 — for signals of escalation.

Fintech And Energy Startups in Africa Captured 64% of Total Amount Raised in 2025 – Report

0
Fund, money cash dollar

In 2025, Africa’s startup funding landscape retained familiar sector rankings even as the overall market structure continued to evolve.

After declining from $3 billion in 2023 to $2.2 billion in 2024, total funding rebounded to $3.2 billion in 2025.

Report by Africa: The Big Deal, revealed that Fintech remained the leading sector by total capital raised but did not significantly expand its base. The sector secured $1.2 billion in 2025, up slightly from $1.1 billion in 2024, across 124 companies, fewer than the previous year.

The five largest fintech fundraisers M-KOPA, Wave, MNT-Halan, Moniepoint, and ValU, accounted for $607 million, representing 52 percent of the sector total, compared with 58 percent in 2024 and 66 percent in 2023.

Equity financing remained the primary funding mechanism for fintech at $685 million, but debt financing played a substantial role at $467 million, helping sustain the sector’s overall performance despite fewer funded companies.

Large facilities such as Wave’s $137 million debt raise and MNT-Halan’s bond issuance significantly influenced sector totals. On the exit front, 2025 recorded 49 exits across all sectors, up from 22 in 2024, with fintech accounting for 19 of those exits.

Energy emerged as the sector where the market’s shape shifted most noticeably. The sector raised $857 million across 50 companies, rebounding sharply from $445 million in 2024 and returning close to its 2023 level of $792 million.

However, concentration intensified, with the five largest energy companies accounting for $701 million, or 82 percent of total sector funding, up from 79 percent in 2024 and 75 percent in 2023.

Debt financing was the primary driver of both growth and concentration in energy, accounting for $611 million, 71 percent of the sector’s total funding. Large deals from d.light ($300 million), Sun King ($156 million), and BURN ($80 million) underscored how a small number of sizeable facilities can significantly shape sector performance.

This dynamic has made energy structurally distinct from other sectors, with large debt-driven transactions disproportionately influencing totals.

Beyond fintech and energy, other sectors recorded lower funding totals but broader participation. Logistics and Transport raised $309 million across 63 companies, with 87 percent of funding coming through equity.

Healthcare secured $211 million across 49 companies, also largely equity-led, although a single major round by LXE Hearing accounted for roughly 47 percent of the sector’s total funding. Agriculture and Food raised $122 million but demonstrated notable breadth with 62 funded companies.

Climate Tech continued to stand out as a cross-sector theme rather than a standalone category, spanning industries such as energy, agriculture, and logistics. In 2025, Climate Tech companies raised $1.2 billion across 149 companies, representing 38 percent of total funding for the year.

This compares with $761 million (34 percent) in 2024 and $1.1 billion (38 percent) in 2023. Participation in Climate Tech has also steadily expanded. The segment accounted for 29 percent of funded companies in 2025, up from 28 percent in 2024 and 26 percent in 2023.

This marks a notable shift from 2021–2022 levels, when Climate Tech represented approximately 18–20 percent of funded startups, positioning it as one of the few investment themes combining large capital inflows with increasing market breadth.

Looking ahead, Africa’s funding landscape is likely to remain shaped by large, structured deals rather than broad-based early-stage expansion.

The continued prominence of debt financing particularly in capital-intensive sectors such as energy and infrastructure, suggests investors are prioritizing scalable, asset-backed models with clearer revenue visibility.

Fintech is expected to maintain its leadership position, but growth may depend more on consolidation, infrastructure expansion, and profitability milestones than on rapid startup proliferation.

Notably, energy and climate-aligned investments are poised to attract sustained institutional interest as governments and development partners intensify efforts around electrification, industrial resilience, and sustainability.

Overall, the market trajectory points toward deeper capital concentration, greater reliance on blended finance structures, and increased investor selectivity trends that could strengthen mature ventures while making capital access more competitive for early-stage startups.

OpenClaw Founder Peter Steinberger Joins OpenAI as Personal Agent Race Intensifies

0

Peter Steinberger, the Austrian developer behind the viral AI assistant now known as OpenClaw, has joined OpenAI, in a move that underscores intensifying competition for top engineering talent in the fast-expanding artificial intelligence industry.

OpenClaw — previously called Clawdbot and later Moltbot — gained rapid attention for branding itself as the “AI that actually does things,” positioning the system as a task-executing assistant capable of managing calendars, booking flights and interacting autonomously with other AI agents. The name was first changed after Anthropic reportedly threatened legal action over its similarity to Claude, before being rebranded again as OpenClaw.

In a blog post announcing his decision, Steinberger said that while he could potentially have turned OpenClaw into a large standalone company, that outcome did not align with his ambitions.

“What I want is to change the world, not build a large company[,] and teaming up with OpenAI is the fastest way to bring this to everyone,” he wrote.

OpenAI CEO Sam Altman said on X that Steinberger will “drive the next generation of personal agents.” Altman added that OpenClaw will transition into a foundation as an open-source project that OpenAI will continue to support.

From Chatbots to Autonomous Agents

Steinberger’s recruitment highlights a broader shift underway in AI development. The first wave of generative AI products centered on conversational interfaces — chatbots that could generate text, code, and images. The emerging phase focuses on “agentic” systems capable of executing multi-step actions across digital environments.

OpenClaw attracted attention because it emphasized execution rather than dialogue. It aimed to integrate directly with software tools, coordinate tasks, and operate semi-autonomously within structured workflows. That approach aligns with a wider industry pivot toward AI systems that can perform operational work, not merely provide suggestions.

OpenAI has been expanding its own capabilities in tool use, memory persistence, and workflow automation. Bringing in an independent builder who rapidly prototyped a viral, execution-oriented assistant suggests an acceleration of that strategy.

AI’s Escalating Talent Wars

The hire is also part of a deepening pattern of talent consolidation in the AI sector. As frontier model development becomes more capital-intensive and infrastructure-heavy, large labs have increasingly recruited founders, researchers, and engineers from startups and rival firms.

Over the past two years, companies including OpenAI, Anthropic, and major technology firms have competed aggressively for specialists in model architecture, reinforcement learning, infrastructure optimization, and agent design. Compensation packages in top-tier AI roles have surged, and independent projects that demonstrate viral traction or technical differentiation are frequently absorbed into larger platforms.

The poaching of standout developers is not confined to research scientists. Product-focused engineers who demonstrate the ability to translate models into usable, scalable tools have become equally valuable. Steinberger’s trajectory — building a consumer-facing agent that quickly captured attention — fits that pattern.

Industry observers note that the pace of recruitment shows little sign of slowing. As AI systems become more capable, the marginal impact of highly skilled individuals can be significant, particularly in areas like agent orchestration, multimodal integration, and enterprise deployment. With generative AI now central to strategic roadmaps across technology firms, talent acquisition has become both defensive and offensive.

OpenAI aims to balance ecosystem optics with strategic consolidation by placing OpenClaw under a foundation structure while integrating its creator internally. Maintaining an open-source presence can help preserve developer goodwill and mitigate criticism over centralization, even as core capabilities migrate into proprietary platforms.

The move also signals a pragmatic calculation for independent developers. Building a standalone AI company requires not only technical differentiation but also access to compute, distribution channels, and regulatory navigation. For some founders, integration into a well-capitalized frontier lab offers faster scale and broader impact.

Steinberger’s appointment arrives at a moment when the AI industry is shifting from model race headlines to application-layer competition. The next frontier is likely to be defined by systems that can plan, coordinate, and execute across digital ecosystems.

That competition is expected not be limited to product features. Some expect it will also hinge on who can assemble and retain the strongest teams.

With demand for high-level AI expertise far exceeding supply, recruitment battles are poised to remain a defining characteristic of the sector.