DD
MM
YYYY

CATEGORIES

PAGES

DD
MM
YYYY

spot_img

CATEGORIES

PAGES

Home Blog Page 30

Justin Sun Settles SEC Crypto Fraud Case for $10m as U.S. Policy Shift Signals Softer Regulatory Climate

By
Samuel Nwite
-
0

Chinese cryptocurrency entrepreneur Justin Sun has agreed to pay $10 million to settle a civil fraud case brought by the U.S. Securities and Exchange Commission, drawing a close to a high-profile enforcement action that underscored Washington’s shifting approach toward the digital asset industry.

The settlement, disclosed Thursday in a letter filed in federal court in Manhattan, still requires approval from Edgardo Ramos. The payment will be made by one of Sun’s affiliated companies. Under the terms of the deal, Sun and the entities named in the lawsuit neither admitted nor denied wrongdoing.

The case, first filed in March 2023, accused Sun and several companies tied to him — including Tron Foundation, BitTorrent Foundation, and Rainberry — of orchestrating a scheme to illegally distribute cryptocurrency tokens and manipulate market activity.

Regulators alleged that Sun generated roughly $31 million in illicit proceeds through what the SEC described as extensive wash trading. According to the complaint, Sun directed employees to carry out hundreds of thousands of trades involving the tronix token between accounts he controlled, creating a misleading impression of genuine investor demand and trading volume.

The agency also accused Sun of secretly paying celebrities to promote the Tronix and bittorrent tokens on social media while failing to disclose the financial arrangements to investors.

Among the personalities cited in the complaint were actress Lindsay Lohan, singers Akon and Ne-Yo, and internet personality Jake Paul. The SEC said their endorsements helped drive retail investor interest in the tokens without revealing that the posts were paid promotions.

Sun welcomed the settlement, describing it as the end of a prolonged legal dispute.

“I am pleased to confirm that the SEC has moved to dismiss all claims against me, Tron Foundation and BitTorrent Foundation,” Sun said in a statement posted on X. “Today’s resolution brings closure.”

The resolution comes at a moment when U.S. cryptocurrency regulation is undergoing a noticeable recalibration. The case was originally pursued during the tenure of former SEC Chair Gary Gensler, whose aggressive enforcement strategy against digital asset firms drew sustained opposition from the crypto industry.

In February 2025, shortly after Donald Trump returned to the White House, the SEC placed the case on hold to explore a negotiated settlement. Trump has repeatedly pledged to make the United States the world’s leading hub for cryptocurrency innovation, a position that has encouraged expectations of lighter-touch regulation compared with the previous administration.

Sun’s growing ties to the U.S. political landscape have also attracted scrutiny. The entrepreneur has emerged as one of the most prominent buyers of the World Liberty Financial cryptocurrency token, a digital asset project in which Trump holds a partial ownership stake.

That connection has fueled criticism among some lawmakers, who argue the case settlement raises questions about regulatory independence.

Elizabeth Warren, the top Democrat on the Senate Banking Committee, condemned the agreement in a sharply worded statement.

“The SEC should not be a lap dog for Trump’s billionaire buddies,” Warren said.

The White House rejected that characterization. Spokeswoman Taylor Rogers said the administration’s policies toward the cryptocurrency industry are designed to strengthen economic competitiveness.

“The President is and always has been motivated solely by what is best for the American people,” Rogers said.

Beyond the legal resolution, the case highlights broader tensions between regulators and the rapidly expanding digital asset sector. Over the past several years, the SEC has pursued multiple enforcement actions targeting what it says are unregistered securities offerings, undisclosed promotions, and market manipulation within cryptocurrency markets.

Industry leaders, however, have argued that the absence of clear legislative frameworks has forced regulators to rely heavily on lawsuits, creating uncertainty for companies and investors alike.

Sun’s settlement removes a major legal overhang for the Tron ecosystem and its associated tokens. Analysts say the outcome may also signal a more pragmatic phase in U.S. crypto regulation, where negotiated settlements and policy reforms increasingly replace the sweeping enforcement actions that characterized the earlier years of the industry’s clash with federal regulators.

A Look At OpenAI’s Newly GPT-5.4 Frontier Model

By
Paul Ugbede Godwin
-
0

OpenAI released GPT-5.4. This is the latest frontier model in the GPT-5 series, described by OpenAI as their “most capable and efficient” version yet, specifically optimized for professional work, knowledge tasks, coding, and agentic workflows.

It unifies advances from previous releases like GPT-5.3-Codex for coding into a single system. For the first time in a general-purpose OpenAI model, it can control a computer like a human — navigating interfaces, clicking, typing, browsing, and working across apps. This enables true AI agents for automation.

1 million token context window (in API/Codex): Supports very long contexts (up to ~922K input / 128K output in some configs), great for complex, multi-step tasks without losing track. Improved efficiency and factuality: Uses significantly fewer tokens (up to 47% less on some tasks), runs faster, hallucinates less (33% fewer false claims vs. GPT-5.2, 18% fewer overall errors), and retains context better during long “thinking” sessions.

In ChatGPT, you can now interrupt the model while it’s thinking and generating and adjust instructions or direction on the fly (rolling out on web/Android now, iOS soon). Enhanced reasoning, coding, and tools: State-of-the-art on benchmarks for professional tasks like spreadsheets, documents, presentations, web research.

It supports “reasoning.effort” levels and excels at agentic planning, execution and verification. GPT-5.4 Pro — Max performance for the hardest tasks. GPT-5.4 Thinking is rolling out to Plus, Team, and Pro users (Enterprise/Edu via admin settings). Pro version for higher tiers. API: Available immediately as gpt-5.4 and gpt-5.4-pro.

Codex: Integrated for coding and agent use. Legacy models like GPT-5.2 Thinking stay accessible for ~3 months until June 2026. Pricing NotesIt’s positioned as premium: higher per-token costs than predecessors; input and output rates reflect frontier status, with multipliers for very long contexts >272K tokens).

But efficiency gains (fewer tokens needed) can offset this for many workflows. This comes just days after GPT-5.3 Instant showing OpenAI’s rapid iteration pace. Early user and dev feedback highlights big jumps in practical agent and autonomy use cases, though some note it’s still evolving amid broader company context.

Native computer use in GPT-5.4 refers to OpenAI’s built-in, state-of-the-art capability that allows the model to directly interact with and control a computer interface — much like a human user would. This is a major advancement toward truly autonomous AI agents, and it’s the first time OpenAI has integrated this natively into a general-purpose frontier model.

The model operates in a visual + action loop:It receives screenshots or screen captures of the current interface. It analyzes what’s on screen using its vision understanding. It decides on the next action and outputs structured commands, such as: Moving and clicking the mouse at specific coordinates.

Typing text or keystrokes. Scrolling, dragging elements, or navigating menus. Your code or harness (the surrounding software) executes those actions in the real environment. It gets the updated screenshot back and repeats — forming a closed loop of observe ? plan ? act ? verify ? correct.

This enables multi-step, real-world workflows without needing pre-built APIs for every tool. GPT-5.4 excels at both: Code-based control — Writing automation scripts using Playwright for browsers Direct low-level control — Issuing raw mouse and keyboard events based purely on visual input.

Developers can steer its behavior through prompts, set custom safety rules; requiring user confirmation for risky actions like deleting files or making payments, and adjust risk tolerance. This beats GPT-5.2’s 47.3% and even surpasses average human performance (72.4%).

It also leads on related agentic benchmarks like WebArena and BrowseComp, showing big gains in reliability for long-horizon tasks. Real-World ExamplesAutomate filling out forms across multiple websites/apps. Pull data from Excel/Google Sheets ? analyze it ? generate a report/presentation. Navigate file systems, open documents, edit them, and save changes.

Debug software by controlling an IDE, running tests, and fixing issues in a loop. Handle repetitive office workflows; data entry, invoice processing, research + summarization across tools. In practice, this turns GPT-5.4 into something closer to a digital employee that can “use” your computer directly, rather than just suggesting steps for you to follow.

It’s designed for professional/enterprise use, with configurable safeguards to prevent misuse. This feature marks a big step in the shift from chat-based AI to action-taking agents — enabling more autonomous, end-to-end automation in knowledge work.

Federal Judge Ordered Trump Administration to Begin Process of Refunding over $130B in Tariffs Collected 

By
Paul Ugbede Godwin
-
0

A federal judge has ordered the Trump administration to begin the process of refunding over $130 billion in tariffs that were collected under President Trump’s emergency-imposed levies.

Judge Richard Eaton of the U.S. Court of International Trade based in Manhattan/New York issued a ruling directing U.S. Customs and Border Protection (CBP) to start calculating and issuing refunds for tariffs deemed illegal by a prior Supreme Court decision in late February 2026.

The Supreme Court (in a 6-3 ruling) struck down the broad “reciprocal” or global tariffs Trump imposed on imports from nearly every country, finding that he overstepped his authority by using the International Emergency Economic Powers Act (IEEPA) — a sanctions law — to unilaterally impose them, as tariff powers belong to Congress.

The tariffs in question were part of Trump’s “Liberation Day” policy last year, leading to collections exceeding $130 billion through mid-December with some estimates suggesting potential refunds up to $175 billion including interest, per the Penn Wharton Budget Model.

Judge Eaton stated that all importers of record whose entries were subject to these IEEPA duties are “entitled to benefit” from the Supreme Court’s ruling — meaning refunds aren’t limited to the over 1,000–2,000 companies including Costco, FedEx, and others that have already sued for repayment.

The process starts with CBP recalculating duties without the invalidated tariffs, though it’s described as complex, potentially drawn-out possibly years, and subject to appeals or stays by the administration. The judge asserted sole jurisdiction over related refund cases and set a follow-up hearing to address implementation details.

The administration is widely expected to appeal or seek delays, with legal experts predicting challenges to slow or limit the refunds. This represents a significant setback for Trump’s trade agenda, as the refunds create a major fiscal liability for the government.

Note that the tariffs were ultimately paid by U.S. importers often passed on to consumers or businesses, so refunds would go to those importers rather than foreign entities or end consumers directly. The ruling has sparked discussion on X with posts highlighting the scale, potential delays, and partisan angles.

Refunds would return significant capital to U.S. importers, including major companies like Costco, FedEx, and thousands of others. This could improve cash flow, enable reinvestment in operations, hiring, or price reductions, and serve as an economic stimulus for affected sectors. Trade groups like the U.S. Chamber of Commerce have called for swift refunds to allow businesses to “reinvest in their operations, employees, and customers.”

Tariffs were largely passed on as higher prices, contributing to inflation Yale Budget Lab estimates added ~$1,400 annually to median household costs in some categories like clothing and electronics. Consumers won’t receive direct refunds, as payments went to importers—not end buyers—creating a one-sided outcome: businesses recover costs, but households absorbed the inflation without reimbursement.

Refunds could lead to repricing of goods, renegotiated contracts, and shifts in import dynamics. However, delays mean lingering uncertainty for small businesses, which may lack resources to pursue claims effectively. The ruling undermines aspects of Trump’s trade agenda, though the administration has shifted to replacement tariffs to maintain revenue.

This creates ongoing volatility in global trade flows. The Treasury faces a massive outflow—$130B+ collected through mid-December 2025, plus interest accruing at roughly $700 million per month or ~$23 million/day during delays, per Cato Institute estimates. Refunds could exceed combined annual spending of departments like Transportation and Justice.

This creates a significant fiscal liability, potentially requiring offsets elsewhere. The administration has resisted quick refunds, seeking delays and may appeal to slow or limit payouts. Refunds involve recalculating duties via U.S. Customs and Border Protection (CBP), handling millions of entries, and likely years of litigation. Judge Eaton asserted sole jurisdiction over cases, with a follow-up hearing around March 6.

Over 2,000 lawsuits are already filed, but not all importers may pursue claims efficiently. The Trump administration is widely anticipated to appeal or seek stays, prolonging uncertainty and adding interest costs borne by taxpayers. This marks another court defeat on trade policy, fueling criticism from opponents and conservative commentary on X framing it as obstructing revenue used to reduce debt.

Discussions on X reflect polarized views—some celebrate it as justice against “illegal” tariffs, others decry it as leftist interference in fiscal gains. While the ruling provides relief to importers and exposes executive overreach on tariffs.

It introduces short-term fiscal strain, prolonged legal battles, and continued trade policy turbulence—without broadly alleviating consumer-level costs from the original levies. The process remains far from automatic or immediate, with appeals likely extending timelines significantly.

Building Africa’s Hardware Future: Embedded Systems & Artificial Intelligence Laboratories for Universities

By
Ndubuisi Ekekwe
-
0

Across the world, the next wave of innovation will not only come from software. It will come from intelligent machines, embedded systems, robotics, and AI integrated into physical devices. From autonomous vehicles and smart factories to intelligent healthcare devices and industrial IoT systems, the future of computing is increasingly embedded and physical.

For Africa to participate meaningfully in this new era, our universities must move beyond teaching only theoretical computing. Students must be able to design, prototype, and test real hardware systems powered by artificial intelligence. This requires modern laboratories that combine embedded systems engineering, semiconductor experimentation, robotics, and AI development. First Atlantic Semiconductors & Microelectronics Ltd (FASMICRO) is helping universities to establish these labs.

FASMICRO – Providing Support To Customers As An Intel Partner

As an Intel Technology Partner and an Altera FPGA partner, FASMICRO works with universities and research institutions across the continent to design and implement state-of-the-art Embedded Systems and Artificial Intelligence Laboratories. These facilities enable students and researchers to build technologies that power modern industries.

An Embedded Systems and AI Laboratory is a specialized facility where students can design and build computer systems that perform dedicated functions within larger mechanical or electrical systems. In practice, this means students can develop systems such as smart sensors, robotics platforms, AI-powered devices, IoT systems, autonomous machines, and edge computing solutions.

The laboratories implemented by FASMICRO typically integrate multiple functional modules.

  • The Embedded Systems Development Module focuses on microcontroller programming, FPGA development, IoT devices, and sensor systems. Here, students learn how software interacts directly with hardware to create intelligent electronic systems.
  • The Artificial Intelligence Development Module enables students to design and deploy AI models, develop AI agents, and experiment with machine learning systems that can operate in real-world environments.
  • The Physical AI and Robotics Module allows students to build robotics systems, autonomous devices, drones, and edge AI machines that combine intelligence with mechanical systems.
  • The PCB Design and Electronics Fabrication Module provides the tools for students to design and prototype electronic circuits using modern CAD systems. This includes PCB design, circuit fabrication, soldering, and hardware debugging, ensuring students understand the full lifecycle of electronic product development.
  • Finally, a Cloud AI Infrastructure Layer provides the computing resources needed for training machine learning models, managing AI workloads, and deploying AI systems.

But infrastructure alone is not enough. A key component of the FASMICRO service is capacity development and academic enablement. In partnership with Tekedia Institute, the company develops complete courseware packages that accompany each laboratory deployment. These include:

  • structured training curricula
    • laboratory manuals
    • hardware design kits
    • FPGA development resources
    • AI experimentation frameworks
    • operational manuals for faculty and lab managers

The goal is not simply to install equipment, but to create a sustainable ecosystem for teaching, research, and innovation.

Through this partnership between Tekedia Institute and First Atlantic Semiconductors & Microelectronics Ltd, universities receive a complete solution: laboratory infrastructure, training programs, academic content, and industry-aligned learning materials. This approach ensures that students graduate with practical engineering capabilities, not just theoretical knowledge.

If you represent a university or research institute, we invite you to contact us at info@fasmicro.com. We would be happy to schedule a Zoom session to discuss how FASMICRO and Tekedia Institute can partner with your institution to empower and educate Africa’s next generation of technology leaders.

Image credit: Fasmicro, 2023

Shorter Message – Share with Your School Administrators

The next wave of global innovation is shifting from pure software to the integration of intelligence into the physical world. From autonomous vehicles and smart factories to AI-powered medical devices, the future of computing is increasingly embedded and physical. For Africa to lead in this era, higher education must evolve beyond theoretical instruction. Students require a practical environment to design, prototype, and test hardware systems powered by Artificial Intelligence.

First Atlantic Semiconductors & Microelectronics Ltd (FASMICRO), an Intel and Altera FPGA partner, is bridging this gap by helping universities establish state-of-the-art Embedded Systems and AI Laboratories. These specialized facilities integrate several critical modules: Embedded Systems Development for FPGA and microcontroller programming; an AI Development Module for deploying machine learning models; Physical AI and Robotics for autonomous machines; and PCB Design and Fabrication for full-cycle hardware prototyping. Supported by a Cloud AI Infrastructure Layer, these labs allow students to transition intelligence from local machines to real-world mechanical systems.

Crucially, FASMICRO understands that infrastructure alone is insufficient. In partnership with the Tekedia Institute, every laboratory deployment includes comprehensive academic enablement. This ecosystem provides universities with structured curricula, laboratory manuals, hardware design kits, and faculty training. This holistic approach ensures that the facility is not just a room full of equipment, but a sustainable center for research and innovation.

By combining cutting-edge hardware with industry-aligned learning materials, FASMICRO and Tekedia Institute are providing a turnkey solution for African universities. This partnership ensures that the next generation of engineers graduates with the practical capabilities needed to build the technologies of tomorrow.

If you represent a university or research institute ready to empower Africa’s future technology leaders, we invite you to contact us at info@fasmicro.com to schedule a consultative Zoom session.

SoftBank Seeks Up to $40bn Bridge Loan to Fund OpenAI Investment Ahead of Expected IPO

By
Samuel Nwite
-
0

SoftBank Group Corp. is in advanced discussions with banks to secure a bridge loan of up to $40 billion, primarily to finance its deepening investment in OpenAI, according to a Bloomberg News report, citing people familiar with the matter.

The facility, which would mark SoftBank’s largest-ever borrowing denominated solely in U.S. dollars, is structured with a 12-month tenor and is being underwritten by four lenders, including JPMorgan Chase & Co. Talks remain ongoing, and terms could still evolve, the report noted. The loan comes on the heels of SoftBank’s completion of a $41 billion investment in OpenAI, finalized in December 2025, which has positioned the Japanese conglomerate as the AI startup’s largest financial backer with an approximately 11% ownership stake.

The commitment was first announced in March 2025, when SoftBank agreed to invest up to $40 billion in a for-profit subsidiary of OpenAI, structured as a combination of direct capital from SoftBank Vision Fund 2 (SVF2) and syndicated co-investments from third-party participants. The funding was executed in two closings. The first, in April 2025, involved $7.5 billion through SVF2. The second closing, completed December 26, 2025, added $22.5 billion from SVF2, plus an oversubscribed $11 billion from co-investors, bringing the total to $41 billion.

This exceeded the initial $40 billion pledge, reflecting strong external demand and SoftBank’s confidence in OpenAI’s trajectory. SoftBank CEO Masayoshi Son has described the OpenAI bet as central to his “all-in” strategy on artificial intelligence, viewing the company as a linchpin in the emerging “Intelligence Revolution.” The investment aligns with Son’s long-term vision of AI as a transformative force, building on SoftBank’s earlier stakes in Arm Holdings (ARM), Alibaba (BABA), and other tech giants.

OpenAI’s latest funding round valued the company at $840 billion post-money, a figure that could climb to $1 trillion in an anticipated initial public offering (IPO) later in 2026, according to Reuters reporting from 2025. To fund the first tranche of the OpenAI investment, SoftBank reportedly relied on an $8 billion bridge loan, as it lacked sufficient cash on hand at the time.

Sources indicate the company may draw on up to $11.5 billion in undrawn margin loans backed by its stake in Arm Holdings to cover portions of the remaining commitment. This reliance on leverage has raised some investor concerns, with  SoftBank’s credit default swaps (CDS) spreads widening in recent months amid broader market volatility in AI-related assets.

S&P Global Ratings has warned that SoftBank’s increasing leverage and concentration of assets around OpenAI could pressure liquidity metrics and credit spreads if market conditions deteriorate. The company’s debt-to-equity ratio has climbed in recent quarters, driven by aggressive AI bets, though strong performance from Arm (whose shares have risen 65% since its 2023 IPO) has provided a buffer.

Analysts at Jefferies noted in a February 2026 report that SoftBank’s finances remain “manageable” but emphasized the need for disciplined capital allocation to avoid overextension.

OpenAI’s IPO preparations are well underway, with the company raising $110 billion in its latest round from investors including SoftBank ($30 billion), Nvidia ($30 billion), and Amazon ($50 billion). The funds are earmarked for expanding AI research, data center infrastructure, and global operations. Reuters reported in January 2026 that OpenAI is targeting a $1 trillion valuation for its public debut, which would make it one of the most valuable listings in history, surpassing Saudi Aramco’s 2019 IPO.

SoftBank’s $41 billion stake, now valued at over $90 billion based on OpenAI’s latest post-money valuation, represents a substantial unrealized gain, underscoring the success of Son’s AI pivot after earlier Vision Fund setbacks. The bridge loan would provide immediate liquidity to fulfill commitments without liquidating other holdings, such as Arm shares, which Son has described as “core” to SoftBank’s portfolio.

The investment, however, has met some controversy. In 2025, MIT Sloan professor Michael Cusumano characterized Nvidia’s parallel $30 billion commitment as “kind of a wash,” noting the circular nature of AI investments where chipmakers fund model developers who, in turn, purchase massive volumes of GPUs. SoftBank’s loan-backed approach adds another layer of financial engineering to this ecosystem, potentially amplifying returns but also risks if AI adoption slows or valuations correct.

Market reaction to the Bloomberg report was muted Tuesday, with SoftBank shares dipping modestly in Tokyo trading amid broader Asia-Pacific weakness driven by Middle East tensions. Investors appear focused on SoftBank’s ability to manage leverage while capitalizing on OpenAI’s growth — especially with the IPO on the horizon.

The $40 billion facility, if completed, would yield the unprecedented scale of capital required to fuel the AI boom, and it underlines SoftBank’s willingness to leverage its balance sheet aggressively. It also highlights the symbiotic relationship between infrastructure providers like Nvidia and model builders like OpenAI, with SoftBank acting as a key financier, bridging the two.

1...293031...8,329Page 30 of 8,329

© Tekedia. All rights reserved.

Term & Privacy

© Tekedia. All rights reserved.

Term & Privacy