OpenAI has held a series of discussions with officials at the U.S. Food and Drug Administration (FDA) over the potential use of artificial intelligence to streamline parts of the drug evaluation process, according to a report by Wired published Wednesday.

The talks mark a significant step toward integrating generative AI tools into one of the most tightly regulated and scientifically rigorous areas of the U.S. federal government.

At the center of the conversation is a project called cderGPT, which appears to be an experimental AI system aimed at assisting the FDA’s Center for Drug Evaluation and Research (CDER). The CDER is the agency’s primary division for overseeing the safety and effectiveness of prescription and over-the-counter medications in the United States.

According to Wired, the meetings have involved not only OpenAI staff but also participants from a newly formed federal agency called the Department of Government Efficiency, or DOGE, an initiative linked to Elon Musk’s broader push to reform how government systems operate through tech innovation. The exact role of DOGE in these conversations remains unclear, but sources familiar with the talks confirmed its presence during several meetings with the FDA.

AI’s Role in Accelerating Regulatory Science

The FDA is reportedly investigating how artificial intelligence could accelerate parts of the drug development timeline, especially during the later stages of review, where volumes of clinical data and scientific documentation must be processed. Drug development in the United States can span more than a decade from pre-clinical stages to FDA approval, with regulatory review often representing one of the more time-intensive phases.

AI could help shrink that window. The FDA has completed its first AI-aided review of a scientific submission for a product. The review was characterized as a milestone in the agency’s modernization efforts.

“The FDA is committed to supporting innovative approaches for the development of medical products by providing an agile, risk-based framework that promotes innovation and ensures the agency’s robust scientific and regulatory standards are met,” said FDA Commissioner Robert Califf. “With the appropriate safeguards in place, artificial intelligence has transformative potential to advance clinical research and accelerate medical product development to improve patient care.”

Jeremy Walsh, a former Department of Defense AI specialist who was recently appointed as the FDA’s AI officer, has been leading the agency’s dialogue with OpenAI. Walsh is also reportedly developing internal guidance on how the FDA could adopt large language models (LLMs) like GPT-4 for regulatory science.

Although the cderGPT project remains largely experimental, people close to the talks told Wired that the conversations have focused on how such a system could assist in data analysis, literature review, and perhaps even drafting parts of regulatory documents—tasks that are often repetitive and time-consuming.

The Promise and Pitfalls of Generative AI in Drug Review

Despite the enthusiasm around AI, many experts warn of the risks. The reliability of large language models, especially in high-stakes fields like medicine, is still under scrutiny. Generative AI models like those from OpenAI have demonstrated incredible capabilities, but they also have real limitations when it comes to hallucinations, transparency, and reproducibility.

This backdrop has created a real need for policy guidance about what datasets are appropriate for training models used in this kind of work, how performance should be measured, and whether the outcomes are interpretable by human reviewers.

According to the Wired report, the FDA is actively investing in understanding these challenges. The agency has launched a fellowship program to train experts on the use of AI in regulatory science and is reportedly experimenting with its own in-house large language models to explore specific applications in areas like precision medicine and post-market surveillance.

A Government-Wide Shift Toward AI?

The involvement of DOGE, a government body reportedly initiated under Elon Musk’s influence, adds another dimension to the story. While DOGE’s official mandate is still largely undefined, it has been described by internal sources as a cross-agency task force aimed at making government more efficient through emerging technologies.

DOGE’s participation in FDA talks with OpenAI suggests a growing federal appetite for AI-led transformation, not just in public services but within scientific agencies that have traditionally operated with high levels of caution and bureaucratic inertia.

Time will tell whether OpenAI’s cderGPT or similar projects become central fixtures of the FDA’s regulatory toolkit. But the willingness of both parties to even explore such a future underlines a meaningful shift in how the U.S. government thinks about innovation, automation, and public health.

VanEck, a major U.S. investment firm, has filed with the SEC to launch the first Binance Coin (BNB) ETF in the U.S., as per a Form S-1 submitted on May 2, 2025. The proposed “VanEck BNB ETF” aims to track BNB’s spot price, the native token of Binance’s BNB Chain, which has a market cap of roughly $84-$88 billion and ranks as the fifth-largest cryptocurrency. The ETF would hold BNB tokens directly and may include a staking feature to earn rewards, pending regulatory approval. This follows VanEck’s earlier filings for Bitcoin, Ethereum, Solana, and Avalanche ETFs, positioning it as a leader in crypto ETF innovation.

Arizona Governor Katie Hobbs vetoed Senate Bill 1025 on May 2, 2025, which would have allowed the state to hold Bitcoin as part of its official reserves, potentially making Arizona the first U.S. state to adopt such a policy. The bill, passed by the Arizona House on April 28 with a 31-25 vote, proposed using seized funds to invest up to 10% of state-managed assets in Bitcoin and create a Digital Assets Strategic Reserve Fund.

Hobbs cited Bitcoin’s volatility and lack of regulatory clarity, calling it an “untested investment” unsuitable for Arizona’s retirement system, which she described as one of the nation’s strongest due to its “sound and informed” investments. She also noted her prior stance against signing bills unrelated to a bipartisan disability funding agreement, resolved on April 24. A companion bill, SB1373, which would allocate up to 10% of Arizona’s rainy-day fund to digital assets, awaits a final vote but faces uncertain prospects given Hobbs’ position.

Arizona joins states like Oklahoma, Montana, South Dakota, and Wyoming, where similar proposals have failed, while North Carolina’s House passed a bill allowing 5% investment in cryptocurrencies. The veto drew criticism from Bitcoin advocates, with some calling it a missed opportunity for Arizona to lead in digital asset adoption.

Coinbase recently released a Bitcoin commercial during the NBA playoffs, emphasizing Bitcoin’s scarcity and its role as a hedge against fiat currency inflation. The ad, aired in May 2025, contrasts the limited supply of Bitcoin—capped at 21 million coins—with the U.S. dollar, which it claims loses value due to excessive printing by the Federal Reserve. Using a visual analogy, it compares a Bitcoin to a quarter and depicts the vast amount of dollars printed annually, highlighting Bitcoin’s fixed supply and the message, “The future of money can’t be printed.”

Tether, the company behind the USDT stablecoin, has announced the development of “Personal Infinite Intelligence,” an open-source AI platform designed to operate on any device without centralized control, API keys, or central failure points. The platform, part of Tether.ai, will feature a modular AI runtime and integrate USDT and Bitcoin payments via WDK, aiming to merge AI with blockchain technology.

CEO Paolo Ardoino highlighted its decentralized approach, emphasizing billions of AI agents in a peer-to-peer network. This move is seen as a step to enhance Tether’s dominance in the stablecoin market, which boasts a $150 billion market cap and $43 billion daily transaction volume.

Michael Saylor’s company, Strategy (formerly MicroStrategy), announced the acquisition of 1,895 Bitcoin for approximately $180.3 million, at an average price of $95,167 per Bitcoin. This purchase adds to Strategy’s substantial Bitcoin holdings, which now stand at 533,539 BTC, acquired for a cumulative $36.1 billion at an average price of $67,656 per coin. The company continues its aggressive

Bitcoin accumulation strategy, funded through equity offerings and other financial instruments, aiming for a $42 billion Bitcoin portfolio by 2027. This move aligns with Saylor’s long-standing view of Bitcoin as a superior store of value, despite market volatility and economic uncertainties.

Google Wallet has integrated zero-knowledge proof (ZKP) technology to enhance privacy in age and identity verification, allowing users to prove they meet age requirements without revealing sensitive personal data like birthdates or full IDs. This system, which leverages cryptographic principles often associated with blockchain, is live across mobile devices and apps using Google’s Digital Credential API.

The dating app Bumble is among the first partners, using digital IDs from Google Wallet for user verification while ZKPs handle age confirmation. The rollout began in the UK, with digital IDs linked to passports, and is expanding to U.S. states like Arkansas, Montana, Puerto Rico, and West Virginia, with plans for 50 more countries. Google also intends to open-source its ZKP tools, potentially setting a new standard for privacy-preserving digital identity.

The Maldives has partnered with Dubai-based MBS Global Investments to develop a $9 billion blockchain and crypto hub in Malé, aiming to diversify its economy beyond tourism and fisheries. The Maldives International Financial Centre (MIFC), spanning 830,000 square meters, is expected to create 16,000 jobs and triple the nation’s $7 billion GDP within four years.

The project, funded through equity and debt with $4-5 billion already secured, offers incentives like zero corporate tax and 100% foreign ownership to attract crypto investors. However, it faces stiff competition from established hubs like Dubai, Singapore, and Hong Kong, which have advanced regulatory frameworks. The initiative is a strategic move to address the Maldives’ debt crisis, with $1.6 billion in repayments due by 2026, but its success hinges on overcoming regulatory and technological challenges.

Nigeria’s Senate has given the green light to two out of four ambitious tax reform bills proposed by President Bola Tinubu, signaling progress in the administration’s effort to reset the country’s battered fiscal system.

But in a clear indication of the political sensitivity surrounding tax policy, lawmakers rejected the proposed increase in Value-Added Tax from 7.5 to 10 percent, choosing instead to offer minor concessions that would allow VAT input claims on fixed assets, overhead costs, and administrative expenses.

The approved bills—the Nigeria Revenue Service Establishment Bill and the Joint Revenue Board Establishment Bill—mark a decisive step toward centralizing and professionalizing tax administration. The former repeals the long-standing Federal Inland Revenue Service (FIRS), paving the way for a new revenue body, while the latter creates a coordinating framework to harmonize tax collection between the federal and subnational governments, a long-standing challenge in Nigeria’s fragmented tax landscape.

Two additional tax bills, the Nigeria Tax Administration Bill and the Nigeria Tax Bill, are scheduled to be debated and potentially passed on Thursday.

The legislative breakthrough comes nearly a year after Tinubu assumed office with a mandate to fix Nigeria’s sluggish economy and expand the tax net. His proposals, transmitted to the National Assembly earlier this year, were part of a broader campaign to reform what many experts have described as an inefficient and poorly enforced tax regime.

“These bills will add immense value to governance and transform how taxes are collected and shared in Nigeria,” Senate President Godswill Akpabio said on the floor, shortly after the upper chamber passed the bills following a rigorous clause-by-clause review. He commended lawmakers for their discipline and signaled that the chamber was ready to work overtime to conclude the pending legislation.

Senate Pushes Back on VAT Hike, Proposes Development Levy Instead

However, Tinubu’s proposed VAT increase, framed by the presidency as necessary to boost government revenue, proved a step too far for senators, given consideration to the rising cost of living and widespread public discontent. In rejecting the hike, the Senate sought to protect already strained Nigerian households from what would have been another inflationary shock.

Instead, the lawmakers devised an alternative strategy: a new 4 percent development levy on certain revenue streams, ring-fenced to continue funding agencies central to education, technology, and national security.

According to the Senate’s resolution, the development levy will be distributed as follows:

• TETFUND – 50%
• Nigerian Education Loan Fund – 15%
• NITDA – 10%
• NASENI – 10%
• National Cybersecurity Fund – 5%
• Defence Security Fund – 10%

Akpabio warned that phasing out funding for these agencies, as initially proposed, would have crippled vital sectors. He noted that Nigeria cannot afford to undermine the institutions that drive innovation, security, and human capital development.

The decision to preserve their funding also reflects broader concerns over youth unemployment, insecurity, and Nigeria’s fragile tech ecosystem.

The Long Walk to The Tax Reforms

The road to Senate approval was filled with friction. The proposed reforms, though technically driven, sparked controversy among interest groups and lawmakers, particularly those wary of how tax changes might affect businesses and ordinary citizens already grappling with high inflation and economic stagnation. Northern leaders were particularly against the reforms, alleging that it would put the region at a significant economic disadvantage.

Deputy Senate President Barau Jibrin shed light on the behind-the-scenes negotiations that resulted in the approval of the bills, revealing that the elders’ committee played a pivotal role in resolving contentious issues. The committee reportedly engaged with religious leaders, regional representatives, and other civil society groups to cool tensions and build consensus.

“It is time to congratulate the entire Senate and, in particular, the committee on finance and the elders’ committee for the wisdom and leadership shown in these bills,” Jibrin said, praising the maturity lawmakers demonstrated in navigating the legislative maze.

Final Passage Expected as House Approves All Four Bills

Both chambers of the National Assembly are now aligned on the direction of fiscal reform, with the House of Representatives already passing all four tax bills. Once the Senate concludes deliberations on the remaining two bills, the final versions will be harmonized and transmitted to President Tinubu for assent.

Many believe the passage of these bills would be one of the most substantial overhauls of Nigeria’s tax architecture in decades. The reform package is expected to simplify compliance, improve tax equity, and widen the tax base—an urgent goal as oil revenues continue to dwindle and public debt soars.

For years, Stripe has quietly underpinned much of the internet’s financial plumbing. On Wednesday, the company turned up the volume.

At its annual Stripe Sessions event, the $50 billion fintech heavyweight rolled out one of its most expansive product overhauls to date, offering a glimpse into what the future of digital finance might look like, from self-learning AI systems to stablecoin-powered transactions and even a surprisingly fast onboarding of tech titan Nvidia.

Stripe’s most striking reveal was an AI foundation model, years in the making, trained on an ocean of payment data—tens of billions of transactions. Emily Glassberg Sands, the company’s head of information, called it a breakthrough.

“Previously, we couldn’t take advantage of our vast data,” she said. “Now, we can.”

This new engine isn’t just about data crunching. It identifies subtle signals that indicate fraudulent behavior—patterns so nuanced that most traditional systems would never catch them. Card testing attacks, a method where fraudsters verify stolen card credentials, have long plagued online businesses. Stripe said its new model boosted the detection of these attacks by 64 percent “practically overnight” for large clients. Its previous systems had already cut such attacks by 80 percent over two years, but the company sees this leap as evidence of what’s possible with more generalized machine learning.

Will Gaybrick, Stripe’s president of product and business, said the model was built through self-supervised learning, allowing it to discover its own features.

“We have found over and over and over again in machine learning, generalized models outperform,” Gaybrick said. “A big part of that is agility. It just performs better and adapts better to changes in fraud patterns.”

Alongside its AI effort, Stripe announced a push into stablecoin-backed multicurrency card products, partnering with startups like Ramp, Squads, and Airtm. The new offering aims to allow businesses across borders to operate in the same currency for the first time—a significant step in reducing the friction that typically comes with international commerce. The move comes only three months after Stripe completed its acquisition of stablecoin platform Bridge, indicating the company’s long-term bet on crypto-backed rails, albeit in a tightly regulated, fiat-pegged form.

Stripe also unveiled its Orchestration feature, which gives businesses the tools to manage and optimize performance across multiple payment providers from a single dashboard, regardless of whether they process payments with Stripe. The offering is designed to give enterprise-scale users more control and resilience, particularly in global operations where reliance on a single provider can be risky.

One of the event’s more unexpected headlines came from Stripe’s relationship with Nvidia. According to Vivek Sharma, Stripe’s head of revenue automation, Nvidia migrated its entire subscriber base to Stripe Billing in just six weeks—a process that typically takes several months. The switch marked the fastest-ever migration to Stripe Billing. Nvidia was already a Stripe Payments customer, but this billing transition demonstrated the growing trust enterprise customers are placing in the fintech’s broader ecosystem.

Stripe also revealed that it now supports 25 new payment methods, including UPI in India and PIX in Brazil, bringing its total global offering to more than 125 payment methods. Klarna, the Swedish buy-now-pay-later giant, will soon be available through Stripe’s consumer payments product, Link, starting this summer.

In a move aimed at physical retailers, Stripe Terminal can now be used with third-party hardware, beginning with Verifone. That shift gives businesses more flexibility at the checkout counter, signaling Stripe’s intent to play a broader role in the point-of-sale market.

Stripe also launched Managed Payments, a merchant-of-record solution that allows companies to expand internationally without dealing directly with tax codes, fraud prevention systems, dispute handling, and fulfillment requirements. The idea is to give businesses a plug-and-play infrastructure to operate globally without having to build in-house compliance and operations teams from scratch.

The company is leaning deeper into automation across the board. Smart Disputes is a new tool that uses AI to automate the process of handling payment disputes, while Stripe Tax has expanded its availability from 57 countries to 102. The service now covers the full tax lifecycle, from monitoring and registering to collecting and filing, making it one of Stripe’s most comprehensive global compliance tools.

Finally, Stripe introduced Global Payouts, which enables businesses to send money to customers, contractors, and other third parties using nothing more than an email address. The simplicity is the point: Stripe wants to make cross-border payments as seamless as sending an invoice.

Perhaps what says the most about Stripe’s evolving role in the tech ecosystem is who’s already using its tools. The company named several AI-focused startups—including OpenAI, Anthropic, Cursor, Perplexity, Windsurf, and Eleven Labs—that rely on Stripe Billing. These are some of the most closely watched players in the new wave of artificial intelligence, and their reliance on Stripe’s infrastructure suggests that the company is increasingly seen as a foundational layer not just for e-commerce, but for innovation itself.

With this slate of announcements, Stripe is making it clear that it no longer sees itself as merely a payments processor. It’s building toward becoming an all-in-one financial operating system for the internet economy—one that sits at the intersection of automation, compliance, global payments, and emerging technologies like AI and stablecoins.