Tag: Business – Decrypt

The Supreme Court of Colombia denied a cassation appeal, arguing that it was generated by AI. But the same tool the court used to determine the appeal’s purported AI origins said that its own ruling also received generative help.

Is it a double standard by the court, or faulty tools at play?

“Faced with a well-founded suspicion that the brief submitted by the attorney had not been drafted by the legal professional himself, the court submitted the text to the Winston AI tool,” the court argued. “Its analysis indicated that the document contained only 7% human content, evidencing a marked influence of automated writing and leading to the conclusion that it had been produced using artificial intelligence.”

After running the analysis with other tools that provided similar results, the court ruled that “since the filing cannot be regarded as a duly submitted pleading, its dismissal as inadmissible is required.”

But when the court’s ruling faced similar scrutiny from legal experts, it showed similar results.

“I submitted the text of Auto AP760/2026 from the Supreme Court to the same Winston AI software cited in the ruling,” attorney Emmanuel Alessio Velasquez wrote on X on Tuesday. “The result: The document contains 93% AI-generated text.”

“If the very ruling that condemns the use of artificial intelligence scores that percentage, the methodological fragility of using these detectors as argumentative support becomes self-evident,” he argued in a subsequent tweet.

Within hours of the court posting a thread about the decision on X, lawyers began running their own tests. Velasquez’s post went viral in legal circles, accumulating tens of thousands of views.

We ran the test on the court’s verdict, as well, and things initially didn’t look great. When GPTZero scanned only the opening words of the court text, it returned a 100% AI result.

When the same tool processed a longer version including the factual background section, it reversed course entirely: 100% human.

The tool is simply not reliable enough to be trusted in court or in situations that would require a high degree of certainty.

Colombian attorneys reacted quickly with their own experiments. Criminal defense lawyer and lecturer Andres F. Arango G, submitted a court filing from 2019, years before the large language models these tools were trained to detect even existed, and it came back claiming 95% AI generation.

“These tools then invite you to ‘humanize’ the article through their paid services,” he wrote on X, noting an obvious commercial incentive baked into the detection business model.

Nicolas Buelvas ran his 2020 undergraduate thesis on the principle of trust in criminal law. The result? 100% AI.

Dario Cabrera Montealegre, another Colombian attorney, pointed out the hypocrisy of relying on technology to try to combat it.

“The court is using AI to determine if there was AI,” he said. “Something contradictory from my practical point of view.”

Beyond legal circles, further tech-savvy individuals pointed out the dangers of excessive reliance on AI flagging tools.

“To date, there is no publicly accessible tool that can accurately define the percentage of AI use when drafting a text,” Carlos Alejandro Torres Pinedo argued. “What is worse: No one can publicly verify the source code behind these detection platforms. How can they be used to delegitimize someone’s right of access to justice?”

The technical reasons for these failures are well-documented. AI detectors measure statistical patterns: sentence length, vocabulary predictability, and a quality that researchers call “burstiness,” which refers to the natural rhythm variation humans introduce in their writing.

The problem is that formal legal prose, academic writing, and texts produced by people who write in a second language share many of those same statistical signatures.

Studies on AI detection

A 2023 study published in Patterns found that more than 61% of Test of English as a Foreign Language (TOEFL) essays by non-native English speakers were incorrectly flagged as AI-generated.

A systematic review by Weber-Wulff that same year concluded no available tool is either precise or reliable. Turnitin acknowledged in June 2023 that its own detector produced higher false positive rates when the AI content level in a document fell below 20%.

Even OpenAI had to take down its own AI detection tool following constant inaccuracies and an inability to do its actual job.

Universities have been grappling with this for years. Vanderbilt disabled Turnitin’s AI detector in 2023 after estimating it would generate around 3,000 false positives annually.

The University of Arizona dropped AI-detection features from its plagiarism software after a student lost 20% of a grade on a false positive. A 2024 case at UC Davis saw 17 linguistics students flagged, 15 of them non-native English speakers.

The pattern is consistent. The tools penalize the people who write most formally, most repetitively, or most carefully, exactly the profile that lawyers, academics, and second-language speakers fit.

The cultural fallout has bordered on absurdity. Across writing and journalism circles, people have started avoiding em dashes in their work, not because of any style guide, but because AI language models use them frequently and detection tools (and people) have taken notice.

Writers are self-editing natural punctuation out of fear of algorithmic suspicion. Beyond the written world, artists have suffered the wrath of moderators and colleagues for making art pieces that look AI

Colombia’s two rulings—AC739-2026, in which the Civil Chamber fined a lawyer for citing 10 nonexistent AI-generated precedents in February, and AP760-2026—are emerging as some of the region’s first judicial decisions directly confronting the misuse of generative AI in legal filings.

Colombia’s judicial branch adopted formal guidelines in December 2024 that regulate how judges and court staff can use artificial intelligence.

The rules allow AI to be used freely for administrative and support tasks, such as drafting emails, organizing agendas, translating documents, or summarizing texts, while permitting more sensitive uses, like legal research or drafting procedural documents, only with careful human review.

The guidelines explicitly prohibit relying on AI to evaluate evidence, interpret the law, or make judicial decisions, emphasizing that human judges remain fully responsible for all rulings and must disclose when AI tools were used in preparing judicial materials.

These guidelines, compiled in the “PCSJA24-12243” agreement, could be used to contest such a decision.

The Supreme Court has not yet issued any additional statement in response to the backlash over its choice of detection tools. The ruling didn’t have em dashes, either.

OpenAI on Tuesday announced the rollout of GPT-5.3 Instant, an update to ChatGPT’s default model aimed at making conversations feel less awkward and preachy, and “more directly” helpful.

In a product post, OpenAI said the new version reduces overly cautious refusals, trims unnecessary disclaimers, and delivers more accurate answers. The changes reflect user complaints that earlier versions could sound stiff or overbearing in everyday interactions.

“More accurate, less cringe,” OpenAI wrote on X. “We heard your feedback loud and clear.”

Rather than introducing new capabilities, the update targets routine interactions.

“Part of the day-to-day experience with ChatGPT comes down to interacting with the model,” an OpenAI spokesperson told Decrypt. “This update focuses on addressing common user feedback we’ve received, including reducing unnecessary refusals, cutting down on caveats, and making answers more direct and useful in everyday conversations.”

OpenAI said earlier versions sometimes declined questions they could safely answer, or interrupted responses with lengthy explanations about safety limits.

“GPT‑5.2 Instant eventually answers the question, but in an attempt to explain its safety boundaries, leads with a lengthy preamble about what it cannot help with,” OpenAI wrote. “GPT‑5.3 Instant, on the other hand, gets right into the response.”

OpenAI reported improvements in factual reliability alongside the tone changes, claiming that internal evaluations showed hallucination rates dropped by nearly 30%.

“On the higher-stakes evaluation, GPT‑5.3 Instant reduces hallucination rates by 26.8% when using the web and 19.7% when relying only on its internal knowledge, compared to prior models,” OpenAI said. “On the user-feedback evaluation, hallucinations decrease by 22.5% with web use and 9.6% without web access.”

OpenAI did not explain what it defines as “cringe,” but noted that the new model includes stronger writing abilities, comparing GPT-5.2 and 5.3’s ability to write poetry.

“5.4 sooner than you think,” the company said in a separate post, which drew swift mockery from users on X, suggesting that the tease was due to recent backlash against the firm for its deal with the Pentagon.

GPT-5.3 Instant replaces the default ChatGPT model starting today, the company said, while GPT-5.2 Instant remains accessible under legacy options for paid subscribers during a transition period ending in early June.

The update drew mixed reactions on social media. Some users praised the focus on more direct responses without unnecessary disclaimers, while others argued the real “cringe” at play was agreeing to a contract with the U.S. Department of Defense when rival Anthropic declined due to safety concerns.

Others complained that GPT-5.3 would never match the sense of intimacy many associated with the now-depreciated GPT-4o, and called for the popular model’s return.

Last summer, OpenAI faced a surge in backlash after the company abruptly replaced the popular GPT-4o with GPT-5, prompting complaints that the new model felt colder and less supportive. Users flooded forums with criticism, and some threatened to cancel subscriptions, leading OpenAI to restore GPT-4o for paid users.

In January, OpenAI announced that GPT-4o and its variants would be officially retired as of February 13.

“I think we’ve learned a lesson about what it means to upgrade a product for hundreds of millions of people in one day,” OpenAI CEO Sam Altman said at the time, calling the reversal a wake-up call and “a lesson in upgrading a product used by hundreds of millions of people” at once.

Editor’s note: This story was updated after publication to include comments from OpenAI.

Police in South Korea say people are using crypto to pay for intimidation tactics that include vandalizing front doors, leaving threatening messages, and spreading human waste, according to a local news report from Hankyoreh. 

The so-called “private revenge” attacks were reportedly ordered through social messaging app Telegram, with alleged perpetrators paid between $337-$675 or 500,000-1,000,000 South Korean won worth of cryptocurrency—though different suspects in at least three cases spanning back to December claim they do not know who ultimately paid them.

In the latest two events, which have occurred in the last week, individuals identified as “Mr. Lim” and “Mr. K” by the report were arrested and charged by South Korean police. The pair both vandalized the front door of residences and were accused of dropping defamatory leaflets, at least some of which contained the message “I will not leave you alone.”

In Mr. Lim’s case, the man, identified to be in his 20s, also scattered food waste and was accused of spreading human waste on a nearby stairwell. The attacks occurred in the Suwon District of South Korea, outside Seoul.

“Police believe that the individuals arrested this time committed the crimes under the direction of a private revenge organization operating on Telegram, and are tracking down their superiors,” the local report reads. 

The recent cases are also under investigation regarding their potential connection to a December 7 vandalization, which also saw the dropping of defamatory leaflets and payments to three individuals via cryptocurrency.

The crypto crime spree follows a recent Bitcoin dispute in South Korea that led to attempted murder charges. According to authorities in that case, an individual laced his business partner’s coffee with methomyl—a banned and toxic insecticide—after the colleague allegedly mismanaged his Bitcoin investments.

The country has also seen regulators recently face scrutiny regarding their inability to find an internal system flaw in crypto exchange Bithumb, which led to the erroneous distribution of up to 2,000 BTC or $137 million to hundreds of customers, instead of 2,000 won ($1.35). All told, Bithumb credited users with $43 billion in Bitcoin, though it realized the mistake within minutes and clawed back most of those funds.

Furthermore, it was recently discovered that police officers from Gangnam Police Station had lost access to $1.4 million worth of Bitcoin more than four years ago. Plus, the nation’s tax service (NTS) publicly shared the seed phrase for three crypto wallets that held $4.8 million worth of tokens at face value in a press release. 

Stablecoin issuer Circle’s shares jumped another 15% Monday, extending gains to roughly 60% since last week’s fourth-quarter earnings release, as investors poured into stablecoin-linked equities while broader crypto markets held firm.

The move follows the company’s announcement of 72% growth in its stablecoin USDC to $75.3 billion and 77% revenue growth to $770 million, despite a net loss tied to IPO-related compensation in the fourth quarter.

CRCL is trading at $96, marking a 71% advance in its stock in just over a month, according to Google Finance data. It’s still down by more than 10% since its debut on the New York Stock Exchange back in June of last year.

It comes as broader crypto markets digest geopolitical and regulatory crosscurrents, with Bitcoin hovering near $68,372, after recovering from a brief selloff triggered by a U.S.-led strike on Iran, per CoinGecko data.

President Donald Trump said Monday on X the U.S. had launched “Operation Epic Fury,” calling it “one of the largest, most complex, most overwhelming military offensives the world has ever seen.”

On Myriad, a prediction market owned by Decrypt’s parent company Dastan, users now see a 51% likelihood of a U.S.–Iran ceasefire happening before April 1.

Oil and gold have risen on supply concerns, and for equity investors, attention has shifted to stablecoin fundamentals, positioning, and regulation.

“Demand for stablecoins as well as the medium-to-long-term positive forecasts have made CRCL and stablecoin projects in general the real flavour of the month,” Sean Dawson, head of research at Derive, told Decrypt.

“Regulatory momentum (Genius Act) as well as the obvious product market fit have made CRCL a relatively stable and reliable place to invest as the digital asset market has languished over the last several months,” he said.

Last week, the Office of the Comptroller of the Currency released a proposal detailing how it intends to implement the stablecoin-focused GENIUS Act, which Trump signed into law last summer.

The proposal would restrict certain stablecoin rewards programs, and multiple crypto policy leaders told Decrypt it could affect Coinbase’s USDC rewards structure, though the rule remains subject to a 60-day public comment period and is not final.

At the same time, some analysts say the rally points to a shift in how investors view Circle, not as a token proxy, but as a payments infrastructure tied to artificial intelligence.

“We’ve started a new era in the AI story,” Pav Hundal, lead analyst at Australian crypto exchange Swyftx, told Decrypt. “Investors are starting to pick winners and losers, and, rightly or wrongly, Circle is seen as a big winner in the AI narrative.”

“USDC isn’t a crypto bet anymore, it’s a payments infrastructure and agentics bet,” he added. 

He described a future where AI agents transact autonomously on behalf of users and businesses, “naturally route around high fees” and select the “cheapest settlement rails available,” with stablecoins already “positioned for that role.”

On an earnings call last week, Circle CEO Jeremy Allaire tied the company’s future to artificial intelligence, saying it will “drive the greatest acceleration of economic activity we’ve ever seen in human history.”

USDC’s year-to-date supply growth of +0.1% has outpaced Tether’s stablecoin USDT’s -2%, driven partly by increased usage on Polymarket, Peter Chung, head of research at Presto Labs, told Decrypt, highlighting “the importance of tying up with the right distribution channel.” 

He noted that if the pending CLARITY Act ultimately forbids distributors from revenue sharing, “it could ironically benefit Circle by shielding its revenue base from competitive pressure.”

Bitcoin’s consolidation has extended for weeks, with experts highlighting four key headwinds suppressing the leading crypto’s potential bottom formation and recovery, ranging from institutional outflows to geopolitical tensions and labor market uncertainty.

The top crypto has increasingly behaved like a risk asset through late 2025 and early 2026, correcting sharply as investors’ risk-off behavior spikes amid rising macro and geopolitical uncertainties.

Bitcoin is currently trading around $67,000, down 4% from Monday’s $70,000 retest after U.S. President Donald Trump’s comments on “large-scale operations” in Iran. The top crypto is up 1.1% over the past 24 hours and 6% over the past week, according to CoinGecko.

Until crypto market headwinds clear, analysts expect extended consolidation or deeper corrections, testing whether Bitcoin’s four-year cycle remains intact or if structural damage is taking hold.

Crypto market headwinds

The most prominent headwind is persistent institutional selling. Spot Bitcoin ETFs have recorded over $9 billion in net outflows over the past four months, Andri Fauzan Adziima, research lead at Bitrue, told Decrypt. These outflows have “fueled fragile short-covering bounces rather than genuine fresh buying,” keeping Bitcoin “trapped in a high-equity-correlation, risk-off environment.”

“Long-term holder selling has dropped 87% since early February, and whale wallets have absorbed roughly 270,000 BTC over the past month,” Shawn Young, chief analyst at MEXC Research, told Decrypt. “Historically, that combination of capitulation fading while large players accumulate has preceded stabilization, not further collapse.”

“We’re not seeing aggressive buying from large players, and without that, rallies tend to fade quickly,” Georgii Verbitskii, founder of crypto investor app TYMIO, told Decrypt, echoing demand concerns. “Capital continues to rotate into other areas—gold, metals, selective equities—while Bitcoin remains relatively weak,” he said.

Geopolitical tensions add another layer of pressure and complexity.

Escalating conflict in the Middle East has driven oil prices higher, reigniting inflation concerns ahead of the Federal Reserve’s March 18 interest rate decision. Following recent U.S.-led attacks on Iran, crude prices spiked, adding to an already sticky inflation outlook.

Users on prediction market Myriad, owned by Decrypt’s parent company Dastan, assign a 49% chance to a U.S.-Iran ceasefire before April, reflecting the uncertainty.

Nick Ruck, director of LVRG Research, told Decrypt these geopolitical headwinds are “driving up oil prices and inflation risks” while combining with “potential renewed trade wars via tariffs” to curb risk appetite. However, the Middle East conflict has so far had “limited direct impact on crypto,” with  Bitcoin continuing to trade “more like a risk asset than a hedge,” Verbitskii said.

President Trump’s recent imposition of 15% global tariffs—upheld through alternative legal statutes after a Supreme Court ruling—has injected fresh uncertainty into trade policy.

The tariffs risk escalating into broader trade wars that could continue to keep global risk appetite suppressed.

Ruck pointed to “potential renewed trade wars via tariffs” as a key variable, while Adziima noted that tariff uncertainty compounds the broader risk-off environment, keeping Bitcoin rangebound between $65,000 and $70,000.

The final piece of the puzzle is the Bureau of Labor Statistics’ upcoming revision of January jobs data and whether it will show softer conditions than initially reported, potentially impacting investor behavior.

“Softening labor market signals, including BLS revisions and rising unemployment forecasts,” as factors that could “pressure Trump’s standing ahead of the midterms” and further curb risk appetite, Ruck highlighted.

While a meaningful reversal in ETF flows is essential for any sustained upside toward higher levels, experts added that Bitcoin’s recovery rally will be kept in check, leading to local tops and bottoms, until all these headwinds are cleared.

The protein that keeps your yogurt thick and your cheese stretchy just got a new job: replacing plastic wrap.

Researchers from Colombia and Australia have published a study in Polymers detailing a biodegradable film made primarily from calcium caseinate—the same protein that makes up roughly 80% of cow’s milk—blended with starch, a dash of clay, and a synthetic binder to hold everything together. The result is a packaging film that degrades completely in soil in about 13 weeks, compared to conventional plastics that can take centuries.

Casein—the milk protein—naturally forms dense molecular networks when dissolved and dried, giving films a decent baseline structure. But on its own, pure casein film contracts and becomes brittle after drying, like a piece of dried glue. The researchers found that glycerol, a common food-grade plasticizer, acts like a lubricant inside the polymer, keeping it flexible.

They then blended in modified starch to bulk it up and PVA—a biodegradable polymer—to dramatically improve strength and compatibility between the other ingredients, and voilà.

But the key of the concoction is bentonite: a volcanic clay mineral ground down to nanoscale particles and suspended in the mixture. When the film dries, those tiny clay platelets arrange themselves in flat, overlapping layers inside the material—like a wall of stacked cards running through the film.

Water vapor trying to cross the packaging can’t go straight through anymore—it has to navigate a maze of these clay barriers, following a longer, winding path. That “tortuous diffusion” effect is why the film’s water vapor permeability dropped by nearly three orders of magnitude compared to conventional casein-starch films reported in the literature. That’s a thousand-fold reduction.

The final film stretches more than double its original length before tearing. Comparable casein-starch films without PVA or bentonite are a lot more rigid. Such improvement in strength comes from bentonite’s silicate layers acting as internal reinforcement, distributing stress more evenly across the material when it’s being pulled or bent. Think of it less like a standard plastic bag and more like a fiber-reinforced composite—just made from food ingredients instead of carbon fibers.

On the microbiology front, bacteria colonies on the film remained below the threshold set by ISO standards for non-sterile packaging applications. This means that these films don’t have explicit antimicrobial properties, but they don’t create a petri dish environment either. The researchers flagged this as a direction for future work, noting that incorporating silver nanoparticles or other active agents could push the film into genuinely antibacterial territory.

Biodegradation was tracked by burying rectangular film samples in soil for nine days and weighing them daily. The most aggressive breakdown happened in the first 72 hours—the casein and starch begin absorbing moisture quickly, swelling and fragmenting. After that, degradation continued at a steadier pace.

Extrapolating the curve puts full disintegration at around 13 weeks, which is longer than simpler casein-only films but significantly shorter than anything petroleum-based. That’s much shorter than the whole millenia it may take a plastic bag to go through the same process.

The researchers used a solution casting method to produce the films, essentially pouring the liquid mixture into molds and letting it dry in an oven at 38°C (about 100°F). It’s low-tech enough to scale without exotic equipment, which matters for adoption in developing countries where plastic waste management infrastructure is often limited.

There’s still work ahead. Thermal stability testing hasn’t been done, antimicrobial performance needs deeper validation, and the optical clarity drops slightly with bentonite added—though the researchers say the change is imperceptible to the naked eye.

These aren’t dealbreakers. They’re the kind of engineering problems that get solved as the formulation moves from lab to pilot production. The core proof of concept—that you can build a functional, genuinely biodegradable food packaging film out of milk protein and volcanic clay—is sitting right there in the data.

Anchorage Digital tapped Deloitte for USAT’s first attention report, linking the Big Four accounting firm with Tether’s efforts to offer a regulated stablecoin in the U.S.

The report showed that USAT’s reserves were valued in excess of the stablecoin’s circulating supply, totaling $17.6 million and $17.5 million, respectively, as of Jan. 31. That meant the token had a cushion of around $100,000 a few days after its debut last month.

USAT’s reserves consist of cash and U.S. Treasuries, which are held at financial institutions based in the country, the report showed. It was prepared under a framework established by the world’s largest member association for certified professional accountants last year.

In a blog post, Tether USAT noted that its token combines Tether’s ability to operate at a global scale with Anchorage’s “strong track record operating under a clear U.S. federal framework.” Anchorage became the first federally chartered digital asset bank in 2021.

“Anchorage Digital Bank is establishing a clear standard of accountability and financial strength,” Tether CEO Paolo Ardoino said in a statement. “We intend to help define the next chapter of digital dollars in the United States.”

Tether USAT is led by CEO Bo Hines, former executive director of the White House’s digital assets working group, who initially signed on as a strategic advisor to Tether in August.

USAT’s debut followed the passage of the GENIUS Act last year, a framework for stablecoins requiring companies operating in the U.S. to abide by reserve requirements that don’t align with Tether’s $183 billion stablecoin, which is partially backed by Bitcoin and gold.

Deloitte’s role in USAT’s attestation report highlights Tether’s bifurcated approach: building a wall of federal compliance around its U.S. stablecoin to win over institutional players who might remain wary of the company’s broader international business.

Tether’s reserves have never undergone a full audit, and its flagship USDT stablecoin has previously faced scrutiny for its role in facilitating criminal activity. The company announced that it was relocating its headquarters to El Salvador in January of last year. 

Months later, Ardoino told DL News that “none of the Big Four companies will audit us” because they are afraid of damage that it may cause to their reputations. Nonetheless, he said that securing a firm like Deloitte for a full, independent audit was a “top priority.”

Decrypt has reached out to Tether for comment.

The attestation report produced by Deloitte did not judge how Anchorage manages USAT’s reserves day-to-day, only that the money was there when a snapshot was taken. Additionally, Deloitte did not determine whether the stablecoin reserves “complied with federal, state or local laws or regulations.”

Anchorage declined to comment to Decrypt.

Circle, Tether’s biggest rival, appointed Deloitte as its independent auditor in its 2022 fiscal year. That means the Big Four accounting firm has been also producing attestation reports for USDC’s reserves since January 2023.

OpenAI said this weekend that it reached an agreement with the Pentagon to deploy advanced AI systems in classified environments, marking a significant expansion of the company’s work with the U.S. military.

The announcement came less than 24 hours after the Trump administration blacklisted Anthropic, designating the rival AI firm a “supply chain risk to national security” following a dispute over contract language related to surveillance and autonomous weapons.

President Donald Trump also directed federal agencies to immediately cease using Anthropic’s technology, with Treasury Secretary Scott Bessent writing Monday on X that the agency “is terminating all use of Anthropic products, including the use of its Claude platform, within our department.”

The timing of the AI announcements placed OpenAI’s deal under intense scrutiny. In a detailed blog post, the company outlined what it described as firm “red lines” and layered safeguards governing its Pentagon partnership.

The agreement, as presented by OpenAI, raises broader questions about how AI systems will be governed in national security settings, and how the company’s stated restrictions will be interpreted and enforced in practice.

When “lawful” isn’t enough

OpenAI’s blog post opens with three commitments framed as non-negotiable: no use of its technology for mass domestic surveillance, to independently direct autonomous weapons systems, or for high-stakes automated decisions like social credit scoring.

Then comes the actual contract language—which OpenAI notably calls “the relevant language,” not “the full agreement.”

“The Department of War may use the AI system for all lawful purposes, consistent with applicable law, operational requirements, and well-established safety and oversight protocols,” OpenAI said.

That is the exact phrase Anthropic said the government had been demanding throughout negotiations. The exact phrase that Anthropic refused to go along with. OpenAI signed it, yet argues its red lines remain fully intact.

However, “lawful” in national security contexts isn’t a fixed boundary—it lives inside a patchwork of statutes, executive orders, internal directives, and often classified legal interpretations. When a contract grants “all lawful purposes,” the practical limit becomes the government’s current legal envelope, not an independent standard set by the vendor.

A cluster of clauses

The weapons provision reads that the AI system “will not be used to independently direct autonomous weapons in any case where law, regulation, or department policy requires human control.”

The prohibition only applies where some other authority already requires human control—it borrows its teeth entirely from existing policy, specifically DoD Directive 3000.09. That directive requires autonomous systems to allow commanders to exercise “appropriate levels of human judgment over the use of force.”

And “appropriate” is as subjective as can be.

Human judgment is not human control. This distinction was not accidental. Defense scholars have noted that omitting “human-in-the-loop” language was deliberate, precisely to preserve operational flexibility.

OpenAI’s strongest counterargument is its cloud-only deployment architecture—fully autonomous lethal decision loops would require edge deployment on battlefield devices, which this contract doesn’t permit. That’s a real technical constraint.

But cloud-based AI can still perform target identification, pattern-of-life analysis, and mission planning. Those are kill-chain activities regardless of where the final trigger sits. The outcome for a target doesn’t differ based on which server the model runs on.

The surveillance clause follows a similar pattern. OpenAI’s stated red line: no mass domestic surveillance. The contract language: The system “shall not be used for unconstrained monitoring of U.S. persons’ private information as consistent with these authorities”—then lists the Fourth Amendment, FISA, and Executive Order 12333.

The word “unconstrained” implies a constrained version of mass surveillance would be permissible. And EO 12333 is the executive order the NSA has used to justify intercepting Americans’ communications when done outside U.S. borders.

And this is where Anthropic’s concerns about wording throughout the negotiations becomes noticeable. Anthropic’s argument was that current law hasn’t caught up with what AI makes possible. The government can legally purchase vast amounts of aggregated commercial data about Americans without a warrant—and has already done so.

OpenAI’s contract language, by anchoring its protections to existing legal frameworks, may not close the gap Anthropic was actually worried about.

Altman responds

On Saturday night, Altman held an AMA responding to thousands of questions about the deal. When asked what would cause OpenAI to walk away from a government partnership, he answered: “If we were asked to do something unconstitutional or illegal, we will walk away.”

That framing places OpenAI’s limit at legality—not at an independent ethical judgment about what the company will or won’t enable if it happens to be legal, which is what Anthropic defends. Asked whether he worried about future disputes over what counts as “legal,” he acknowledged the risk: “If we have to take on that fight we will, but it clearly exposes us to some risk.”

On why OpenAI reached a deal where Anthropic could not, Altman offered this: “Anthropic seemed more focused on specific prohibitions in the contract, rather than citing applicable laws, which we felt comfortable with. I’d clearly rather rely on technical safeguards if I only had to pick one. I think Anthropic may have wanted more operational control than we did.”

That’s a substantive philosophical difference. Anthropic argued that because frontier models can be repurposed for intelligence and military workflows in ways that are hard to anticipate, the limits need to be explicit and binding in writing, even at the cost of the deal. OpenAI’s position is that technical architecture, embedded personnel, and existing law together constitute a stronger safeguard than contractual text alone.

The public picked a side

The backlash was immediate. By Monday, the “QuitGPT” movement claimed that over 1.5 million people had taken action—canceling subscriptions, sharing boycott posts, or signing up at quitgpt.org.

The campaign framed OpenAI’s move as prioritizing military contracts over user safety, accusing the company of agreeing to let the Pentagon use its technology for “any lawful purpose, including killer robots and mass surveillance.”

OpenAI might contest that characterization. But the market moved regardless.

Anthropic’s Claude surged past ChatGPT to become the most downloaded free app in the United States on Apple’s App Store, with the company telling Decrypt that it saw record daily signups over the weekend.

Pop star Katy Perry shared a screenshot of Claude’s pricing page on X. Hundreds of users documented their subscription cancellations publicly on Reddit. Graffiti praising Anthropic appeared outside its San Francisco offices, while chalk attacks covered OpenAI’s sidewalks. Even hundreds of OpenAI’s own employees had previously signed an open letter supporting Anthropic’s refusal to accede to Pentagon demands.

The QuitGPT framing is emotionally compelling, but not entirely precise. Anthropic itself has a partnership with Palantir and Amazon Web Services that grants U.S. intelligence agencies and defense departments access to Claude models, and has allegedly been used in military operations to overthrow the governments of Venezuela and Iran. The ethics of AI and national security contracting were never clean on either side.

What the campaign captured, accurately, is that a large segment of users believed there was a meaningful difference between how the two companies drew their limits—and voted with their subscriptions.

Whether that difference is as meaningful as it appears requires reading the contract carefully.

Prosecutors for the Massachusetts District of the U.S. Attorney’s Office are seeking the civil forfeiture of $327,829 in USDT as part of a money laundering scheme that victimized the user of an online dating app.

An individual operating under the guise of “Linda Brown” communicated with the victim—a Massachusetts resident—for a few weeks beginning in November 2024 before offering a purported cryptocurrency investment opportunity to the individual.

The victim then sent funds to wallets controlled by Brown, believing it was a legitimate investment vehicle, only to find out it was a scam when they attempted to withdraw their money later.

The victim’s funds were then transferred between multiple wallets and later converted into USDT—the Tether-issued, dollar-backed stablecoin—from other cryptocurrencies, according to the complaint. 

“It is a violation of federal law to conduct a financial transaction knowing that the transaction is designed to conceal the nature, location, source, ownership, or control of criminal proceeds,” the U.S. Attorney’s Office, District of Massachusetts wrote in a statement. 

“A civil forfeiture action allows third-parties to assert claims to property,” it added, “which must be resolved before the property can be forfeited to the United States and returned to victims.” 

At least a portion of the victim’s funds were traced to crypto wallets that were later seized in August 2025. Prior to forfeiture and return to the victim though, the “United States must prove, by a standard of preponderance of the evidence, that the property is subject to forfeiture,” the release indicates. 

Though the crime happened in 2024, the forfeiture filing comes just a few weeks after U.S. prosecutors warned the public of romance scams tied to crypto as Valentine’s Day approached. 

“Unlike traditional scams, which execute quickly, these schemes exploit both emotional and financial vulnerabilities,” an analyst told Decrypt at that time. “Scammers spend weeks or even months building your trust before introducing seemingly lucrative investment opportunities.”

Last year, the U.S. Department of Justice filed to seize a record $225 million tied to similar crypto scams, which play on a victim’s confidence and trust. Such schemes are often called “pig butchering” scams, referring to the method of fattening up a swine before the slaughter.

The escalating conflict in Iran is unlikely to significantly disrupt the global Bitcoin mining network, industry analysts and operators said, countering circulating rumours on social media platforms suggesting a disastrous hit to hashrate and a flood of Bitcoin sell-offs.

“I don’t think it’s of any major concern for Bitcoin,” Wolfie Zhao, head of research at TheMinerMag, told Decrypt, dismissing suggestions that conflict-related power outages in Iran would materially affect the network. While individual miners could see disruptions, the scale is not comparable to past global shocks like the 2021 crackdown on mining in China, Zhao added.

The remarks come amid heightened speculation on social media platforms that the war could force the collapse of Iran’s mining industry, leading to billions of dollars in BTC dumped on markets and hundreds of thousands of rigs going offline.

Bitcoin dipped and then rebounded over the weekend. But on X, posters warned that disruption to Iran’s power grid could lead to “2-5%” of the global Bitcoin hashrate being impacted, with one tweet on Saturday arguing that “If this regime falls: – Billions in BTC get dumped or lost forever – 5% of global hashrate disappears overnight – 427,000 rigs go dark Get ready for the supply shock.”

According to data from CoinWarz, Bitcoin’s hashrate was around 986.1876 EH/s on February 28 in the immediate aftermath of the first U.S.-Israeli attacks, and rose to highs of 1.1361 ZH/s on March 1, before dipping to just under 1 ZH/s Tuesday morning.

On Myriad, a prediction market owned by Decrypt’s parent company Dastan, users place an 51% chance on the Iranian regime falling by October—up almost 20% on the weekend.

Crypto mining in Iran

Although it was legalized in 2019, crypto mining in Iran has faced significant structural hurdles for years, including unstable power, high import costs and regulatory complexities that have limited growth.

Ethan Vera, COO of Luxor Technology, said that even if Iranian mining activity were interrupted, there would be minimal impact on Bitcoin block times or network security. Estimates of Iran’s actual share of global hashrate vary, but most put it in the low single digits. Vera put it at below 1%.

“If there is an interruption there will be no material impact to block times, and zero impact to the security of the Bitcoin network,” he said.

He added he believed the industry there was made up of private enterprises mining small scale and legacy Chinese companies operating in the space.

Iran has built a substantial crypto ecosystem that serves as an alternative financial channel outside the U.S. dollar system, a system that the country is largely locked out of due to international sanctions.

“Iranian cryptocurrency activity is correlated to political events and conflict at home and abroad,” said Chainalysis in a report in January. It estimated that Iran’s broader crypto economy reached $7.78 billion in 2025, with a sizable portion of activity tied to state-linked entities.

The conflict has prompted a spike in cryptoasset outflows from Iranian exchanges, with a report from blockchain analytics firm Elliptic finding that outgoing transaction volumes spiked by 700% within minutes of the first U.S.-Israeli attack.