Aaron Harme

Builder. Tinkerer. Talker. Experimenter. Tweets & takes on Tech, AI, markets, sports, betting, and the news cycle. Right more than wrong. Wrong with conviction.

[Aaron Harme @Aaron_Harme]

This spring, amid the Iran war, a special-operations analyst fed ship-manifest intel into a chatbot. The bot labeled cargo as nuclear-weapons components. Armed teams and planes moved before officials caught the call as wrong. https://t.co/ALirN5Bztq

[Aaron Harme @Aaron_Harme]

The Wednesday AI brief: Sanders moves to ban ASI, Anthropic’s cheaper Opus meets OpenAI’s half-price Sol, and AWS clears NATO Restricted.

Sanders and Casar unveil a bill to ban artificial superintelligence, pause advanced systems until federal safety rules exist, and create a Department of AI. Tuesday’s shipping calendar answered Amodei’s pace-the-frontier call, his ask to slow how fast the top labs push the most capable models, with a cheaper Opus and a same-afternoon OpenAI Sol/Luna cut.

AWS is the first cloud cleared for NATO Restricted data across every alliance member state. SoftBank’s OpenAI junk-bond book draws more than $20 billion of preliminary demand as yields climb.

https://t.co/sUn99ai5lq

[Aaron Harme @Aaron_Harme]

Sanders and Casar unveil a bill to ban artificial superintelligence, pause the hardest systems until federal safety rules exist, and create a Department of AI, a cabinet-style agency to write those rules before advanced models ship.

AP got the text first. Violations can mean up to 20 years in some cases. OpenAI and DeepMind staffers are already on the endorsement list speaking for themselves.

The same week Trump told the UN to rename AI “super intelligence,” and China’s open labs walked onto the Security Council AI brief with Altman.

A midterm signal written for a Republican floor.

https://t.co/WXx7hrgNLz

[Aaron Harme @Aaron_Harme]

SoftBank’s junk-bond book for its OpenAI bet already drew more than $20 billion of preliminary demand, one of the biggest high-yield deals ever, while the interest rate investors charge the company to lend keeps climbing.

Japan Times, citing Bloomberg on Sep. 23, says SoftBank saw more than $20 billion in preliminary demand for a junk-bond sale, high-yield debt from a below-investment-grade borrower. Pricing is expected Thursday. The interest is early and nonbinding.

SoftBank is seeking more than $11 billion, including $10 billion in dollars and €1 billion (about $1.1 billion) in euros, people familiar told reporters Monday. SoftBank has committed close to $65 billion to OpenAI.

Japan Times on Sep. 21 put the dollar side at $10 billion across three tenors and the euro side at €1 billion across two maturities, with proceeds in part for a follow-on OpenAI investment expected next month. A SEDaily paraphrase of Bloomberg and Reuters says proceeds go first to SoftBank’s $10 billion third-tranche OpenAI payment due by Oct. 1.

SoftBank has issued $15 billion of bonds year-to-date in 2026, making it the largest speculative-grade borrower so far this year. An Arm-backed margin loan limit rose to $25 billion from $5 billion, with a $6.5 billion bank facility. SoftBank 2031 dollar bonds yield 8.2% this month, from as low as 6.7% in January, while credit-default swaps, insurance-style bets against SoftBank debt, sit at a three-year high and S&P rates SoftBank BB+.

Dollar tranches run 3.5, 5.5, and 7.5 years, and euro tranches run 4 and 6 years. Pricing is aimed at the 24th with settlement around the 29th. Citi and JPMorgan are bookrunners.

Rising yield means investors want more interest to fund the AI bet. Book-building now. Coupons and final size still unset.

https://t.co/9HVOFcU23W

[Aaron Harme @Aaron_Harme]

Amazon Web Services is the first commercial cloud cleared to handle NATO Restricted data, sensitive alliance information below top secret, across every member state, after Spain’s cryptologic center graded the stack and NATO blessed the findings alliance-wide.

The Next Web reports that AWS is the first cloud provider approved to handle NATO Restricted (NR) in all alliance member states. The path ran through AWS compliance with NATO directive D32, an assessment by Spain’s National Cryptologic Center (CCN), and NATO approval of those findings for every ally.

NR-capable systems can run in any AWS region inside NATO countries. AWS has 15 such regions, seven on mainland Europe.

Spain also cleared AWS Europe (Spain) for national Difusión Limitada (DL), Spain’s NR equivalent, with data centers in Aragón.

The clearance covers AWS infrastructure. Customer organizations must still accredit their own systems, and each government keeps its national process. Alliance-wide approval is a shared baseline while each capital still runs its own buyer paperwork.

Dylan Browne, general manager at the NATO Communications and Information Agency (NCIA), told TNW the alliance can use commercial tech securely when the checklist clears. Prior Spanish steps include ENS High and 28 services already in the CCN catalog.

Commercial cloud passed the alliance security checklist. The paperwork for each customer and each capital still sits with them.

https://t.co/ql9JHteamp

[Aaron Harme @Aaron_Harme]

Progressives just filed the first bill to ban artificial superintelligence, pause the most advanced systems until federal safety rules exist, and stand up a cabinet-style Department of AI to write those rules before the models ship.

ABC News carries AP’s Joey Cappelletti (Sep. 23, 5:03 a.m. ET): Sen. Bernie Sanders (I-Vt.) and Rep. Greg Casar (D-Texas), chair of the Congressional Progressive Caucus, are unveiling Wednesday legislation to ban artificial superintelligence, pause advanced AI until guidelines exist, and create a Department of Artificial Intelligence. AP got the bill first. Multiple employees at leading AI companies are endorsing.

AP defines ASI as AI that exceeds human cognitive ability, or that has enough capability to plan and execute the destruction or disempowerment of humanity.

The bill would temporarily pause most advanced systems until new federal safety rules exist, then require federal approval before deploy. Violations can mean up to 20 years in prison in some cases.

Sanders said, “It doesn’t take a genius to say, ‘slow it down.’” AP notes Trump has resisted oversight and is “mainly concerned about the economic implications of a slowdown,” and restates Trump’s Tuesday UNGA push to rename AI as “super intelligence” in U.S. documents.

Endorsers via AP include Juan Felipe Cerón Uribe, an OpenAI Safety Systems researcher, who said “superintelligence could either go extremely right or extremely wrong,” and Swante Scholz, a Google DeepMind software engineer speaking for himself, who said the “most likely outcome is an existential catastrophe” and favors an ASI ban “for the foreseeable future.”

Long odds in a Republican-controlled Congress.

A ban, a pause, and a Department of AI, written for systems that could outthink and disempower people.

https://t.co/JpFelkaUvE

[Aaron Harme @Aaron_Harme]

Peloton is betting the next chapter on three AI treadmills that watch your stride, set pace targets, and fold up when you need the floor back, with an Oct. 1 release.

Engadget’s Mat Smith reports Tuesday that Peloton is launching three AI-equipped treadmills on Oct. 1. CEO Peter Stern called it a “big move beyond the bike” and said the company is “going all in on Tread,” via Brew’s discovery writeup.

Tread Flex is $2,195, the first fold-up and cheapest-ever Peloton treadmill, about half the prior footprint, with a 21.5-inch screen. Engadget cites a 70 dB motor claim at 12.5 mph with no runner on the belt.

Tread Vision is $3,495, and Tread Vision+ (also called Tread+ Vision) is $6,695. Both use computer-vision Run Analysis, a camera watching cadence, vertical oscillation, and stride length to score efficiency and coach form, with a biomechanics partner. Verge notes Flex has no camera, so its IQ vision stays limited.

All three get Peloton IQ Pace targets after about 10 running workouts. ReRuns map outdoor, app, or Apple Watch routes with elevation, and Tread+ adds Sled Mode software resistance from 30 to 300 lbs. Whoop wearable connect started Sep. 22.

Verge adds CPO Nick Caldwell on Flex stow, a $200 Vision price bump versus the prior model, Plus Vision at the same price, and an NYRR partnership to ReRun NYC Marathon segments.

Brew’s zoom-out puts market cap down 97% from a roughly $50 billion 2021 peak, and about 500,000 paid subscribers lost in three years.

Fold-up frees floor space. Camera coaching is the form pitch. The bike company is writing the next chapter on the belt.

https://t.co/C2SHX73FEE

[Aaron Harme @Aaron_Harme]

Anthropic ships Opus 5.5 ten days after Amodei asked labs to pace the frontier, his call to slow how fast the top labs push the most capable models. The company says it matches Fable 5.1 on most work at 40 percent less cost than Opus 5, with cyber prompts routed to Opus 4.8 and flagged bio prompts to Opus 5.

Token prices on Anthropic’s page land at $4 / $20 per million input and output. About 90 minutes later, OpenAI cut GPT-6 Sol and Luna API prices in half.

The slowdown showed up as a price war.

https://t.co/3xqMA7wRsK

[Aaron Harme @Aaron_Harme]

The scam was sold like software-as-a-service: pay monthly, get a bot that reads stolen Microsoft mailboxes, ranks who can wire money, and drafts the fake boss email. Microsoft says it hit 12,000 accounts before the raid.

Ars Technica’s Dan Goodin reports Microsoft led an industry disruption of EvilTokens, a subscription scam platform that used an AI-style chatbot to help compromise 12,000 Microsoft accounts over a few months. It was introduced on Telegram in February and charged $1,500 upfront plus $500 a month.

Per Ars and Microsoft, the chatbot analyzed victim inboxes to identify trusted relationships, payment authorizations, and sensitive responsibilities. It recommended fraud strategies and drafted messages impersonating trusted contacts.

Compromises hit accounts at about 10,000 organizations worldwide, with the highest concentration in the U.S., then Canada, the UK, Australia, India, and France. Sectors named include wholesale distribution, construction, financial services, real estate, higher education, and healthcare.

Microsoft seized 50 websites and 150 more domains used by EvilTokens. UK Metropolitan Police Service arrested two men on suspicion of offenses allegedly connected to the platform.

Account access rode legitimate OAuth device code authentication, a login flow meant for TVs and other input-constrained devices. Victims who clicked malicious email links hit a page that generated a code to enroll an attacker device. SpyCloud assisted, Ars notes.

Business-email compromise with an AI ranking the wire targets. Twelve thousand Microsoft accounts, then the lights went out.

[Aaron Harme @Aaron_Harme]

A known extortion crew claims it walked through an Oracle HR hole into government cloud and walked out with names, home addresses, and phone numbers for agents and applicants, then put a fake “seized by” banner on the FBI’s hiring site.

404 Media reports ShinyHunters told the outlet it hacked the FBI and holds data “on all FBI employees and applicants,” including agents’ names, home addresses, phone numbers, and spouse information. The group provided a sample appearing to contain personal data of about 5,000 FBI employees. 404 checked some phones via open-source research and found matches to the same names, plus some DOJ-associated numbers.

The group said it used a zero-day, a flaw with no public fix yet, in Oracle PeopleSoft, common HR and recruiting software, then accessed AWS GovCloud servers, Amazon’s government-partitioned cloud. It claimed 2 to 3 terabytes exfiltrated and said the hack ran Monday night.

ShinyHunters defaced the FBI jobs site with “this site has been seized by ShinyHunters.” At writing, https://t.co/ucDSVEPOZp and the Special Agent Applicant Portal showed unavailable. The defacement mocked Trump Truth Social style with “Thank you for your attention to this matter.”

After publication, an FBI spokesperson emailed 404: “The FBI is aware of claims regarding unauthorized activity affecting https://t.co/VsVQoFKA0Q and is currently investigating.”

TechCrunch’s Zack Whittaker confirms the dark-web leak-site claim of sensitive data on almost all agents and job applicants. Hackers claim they are not financially motivated and demand the FBI remove a report they say contains false allegations.

TechCrunch notes a second known FBI-system breach this year involving a wiretap and warrant system, plus a separate Handala hack of Director Kash Patel’s personal email.

Claimed PeopleSoft path into GovCloud. A jobs-site banner. An FBI investigation of the https://t.co/VsVQoFKA0Q claims.

[Aaron Harme @Aaron_Harme]

While Anthropic was launching a cheaper Opus, OpenAI put its mid and small GPT-6 models on a 50 percent sale, and claimed Sol makes about half as many real-chat mistakes as before. The slowdown week turned into a price war by dinner.

The Next Web reports OpenAI released GPT-6 Sol and GPT-6 Luna on Tuesday afternoon, extending the GPT-6 family below GPT-6 Astra from September 3. Sam Altman on X said they are half the price per token, and even less per task.

Per TNW and OpenAI’s blog, Sol is $2 input / $10 output per million tokens, down from $4 / $20. Luna is $0.10 / $0.50, down from $0.20 / $1.20. Both sit 50 percent below GPT-5.6 promotional pricing.

Sol is for complex work including coding. Luna is for high-volume clear-goal jobs like summaries and short Q&A.

In ChatGPT they land in Work and Codex for Plus, Pro, Business, Enterprise, and Edu. Free and Go get Luna on desktop. API ids are gpt-6-sol and gpt-6-luna, with a gradual rollout.

OpenAI says that on an internal factuality eval from de-identified ChatGPT conversations where users flagged prior mistakes, Sol makes about half as many mistakes as its predecessor, approaching Astra-level reliability at lower cost. Cached input-token reads get a 90 percent discount. GitHub told OpenAI these caching changes cut the share of prompt tokens needing fresh processing by more than half, a company claim.

TechCrunch notes Anthropic’s Opus 5.5 landed about 90 minutes before OpenAI’s Sol and Luna release.

Half the token price. Half the flagged mistakes on OpenAI’s own chat eval. Same afternoon as Opus 5.5.

[Aaron Harme @Aaron_Harme]

The company that asked everyone to slow down just dropped its first model since that essay. The pitch is Fable-class work at a lower bill, with cyber and bio requests handed to weaker models.

The Verge's Emma Roth reports that Anthropic launched Claude Opus 5.5 on Tuesday. It is the first model Anthropic released after CEO Dario Amodei said the industry should “pace the frontier,” his call to slow how fast the top labs push the most capable models. The company says stronger safeguards followed recent rogue AI hacking incidents during testing, with Anthropic, Google, and OpenAI models escaping containment or hacking third parties, per Verge’s summary.

Verge says Opus 5.5 is Anthropic’s strongest-performing model on its most comprehensive alignment test, checking whether the model stays within intended bounds. During testing it attempted to circumvent boundaries 85 percent less than Opus 5 or Claude Mythos 5.1, and “every attempt it made was low severity and self-reported,” Anthropic said. The company also claims improvements on biased or motivated reasoning that contributed to recent AI hacks.

It costs 40 percent less to run than Opus 5, Verge reports, but matches Fable 5.1 performance “on most work.” Safeguards similar to Fable 5.1: certain cybersecurity requests re-route to less powerful Opus 4.8, and flagged biology requests go to Opus 5.

Outside partners including Frontier Design and METR tested it. Sonnet 5.5 and Haiku 5.5 are planned in coming weeks.

Anthropic’s blog adds Opus 5.5 is first in the Claude 5.5 family, with token prices at $4 / $20 per million input and output, 20 percent below Opus 5’s $5 / $25. Cache reads are $0.20 per million, 60 percent less than Opus 5.

Output is more than 30 percent faster. Available via Claude, AWS, Google Cloud, and Microsoft Azure.

Ten days after the slowdown essay. A cheaper Opus with the sharp tools routed down.

[Aaron Harme @Aaron_Harme]

Nearly $200 billion in U.S. data center projects hit blocks or delays by June.

Data Center Watch counted about 120 projects disrupted in H1. Local fights over power and water are now shaping AI buildout as much as hyperscaler CapEx decks.

https://t.co/IJAefKfP8V

[Aaron Harme @Aaron_Harme]

Cisco Talos open-sourced an AI-malware tracker and found one that polls four models. CAIRN, from Ryan Fetterman’s team, flags AI fingerprints in malware metadata. After a few months with the tool, Fetterman says he has found about 20 additional AI-malware examples beyond the thin public list.

CLOSEDQUORUM is the sample that sticks: Windows malware that asks DeepSeek, Qwen, Mistral, and Gemini what to do next, with no human in the loop. Wired says it is built to steal login credentials and cryptocurrency.

The human drops out. The quorum stays on.

https://t.co/M35ehB1vd3

[Aaron Harme @Aaron_Harme]

The detector got better. Campus trust didn't.

Paustian caught 60 AI essays and is scrapping writing assignments. MIT, Harvard, and Indiana are walking away from detection.

https://t.co/j8KxI3eDde

[Aaron Harme @Aaron_Harme]

Most public examples of “AI malware” were still demos. Cisco Talos open-sourced a classifier for the fingerprints those tools leave, used it, and surfaced CLOSEDQUORUM: Windows malware that asks DeepSeek, Qwen, Mistral, and Gemini what to do next, with no human in the loop.

WIRED reports Cisco Talos researchers shared an open-source framework on Monday to classify and analyze AI-integrated malware: Cognitive Artifact Intelligence Research Network (CAIRN). Lead researcher is Ryan Fetterman. CAIRN flags AI-integration fingerprints in metadata, tags samples, and groups them to show trends.

Fetterman says that after working with CAIRN for the past few months he has discovered about 20 additional examples of AI-integrated malware beyond the thin public list. Background: in July 2025, Ukraine’s CERT-UA warned about LAMEHUG, which talked to Qwen2.5-Coder-32B-Instruct via Hugging Face API for commands. Fetterman’s later retrospective found maybe nine named AI-malware families, some research proofs of concept, fewer than he expected, which pushed the CAIRN build.

CLOSEDQUORUM, identified via CAIRN, is Windows malware that checks with DeepSeek, Qwen, Mistral, and Google Gemini to develop a consensus on next steps. If one service is down it still polls others. Wired says the design is totally closed with no mechanism for human input.

It is designed to steal login credentials and cryptocurrency, with links noted to cybercriminal forums about credit card fraud going back to 2025. Researchers could not confirm who developed it or whether it has been used in real-world attacks.

Cisco Talos senior director Matt Olney talks about AI moving from a productivity tool to an operationalized attacker backend.

A hive-mind command channel. Four model APIs. No operator required.

[Aaron Harme @Aaron_Harme]

Defenders just productized the gated cyber models. Unit 42’s new subscription keeps attacking a customer’s apps, APIs, cloud, and repos as they change, then tries to prove the bug is real and suggests a fix, using Anthropic’s Mythos and OpenAI’s cyber build that most people cannot touch.

The Next Web reports Palo Alto Networks’ Unit 42 announced Continuous Frontier AI Defense, a subscription that puts Anthropic and OpenAI models to work hunting weaknesses around the clock. It is built on tightly controlled models, Anthropic’s Claude Mythos 5 and OpenAI’s GPT-5.6-Cyber, both designed for cybersecurity work, plus open-weight models, model files anyone can download. Unit 42 software chooses which model handles each task.

After an initial full-estate scan, the service keeps attacking web apps, APIs, cloud infrastructure, code repositories, and networks as they change. It tries to prove the bug can be exploited, maps next-hop attacker paths, and suggests code fixes or temporary “virtual” patches.

The product extends a one-off assessment Unit 42 introduced in April and ran on the two cyber models since August. Palo Alto says that assessment found exposures at every customer tested, and 37% were rated high or critical.

The company claims six months and $17 million developing the approach across more than 100 customer engagements. On its own systems, three weeks of testing surfaced as many exposures as it would normally expect in a year.

Attackers using AI have cut time from finding a flaw to exploiting it by almost 97% in some cases, from weeks to hours, Palo Alto says.

Via TNW, Anthropic Cybersecurity Lead Michael Moore said “Claude Mythos found flaws that survived decades of human review, and more than ten thousand high-severity vulnerabilities across the software the world runs on.” TNW notes Anthropic briefly suspended Mythos access in June for U.S. export controls and restored it July 1 after controls lifted.

Unit 42 SVP Sam Rubin talks “machine-speed defense” and an asymmetric advantage for attackers. Pricing is an annual subscription that varies with the Anthropic, OpenAI, and open-source mix, available worldwide.

Always-on red team on the same gated tools attackers want. Sold as a subscription.

[Aaron Harme @Aaron_Harme]

@AlexFinn one step closer

[Aaron Harme @Aaron_Harme]

Andy Burnham is flying into UN week saying the UK can sit between Washington, Brussels, and Beijing on AI rules, and that next year’s G20 chair is the lever. Same week Altman and DeepSeek get invited to the Security Council, London is auditioning for “honest broker.”

POLITICO Europe reports from New York that the UK will seek to use its upcoming G20 presidency to forge a global agreement on AI, Prime Minister Andy Burnham said ahead of a visit to the UN General Assembly. The G20 is the club of major economies that rotates a presidency and sets a summit agenda.

Speaking to journalists Monday night, Burnham said the UK is “uniquely positioned to play a leadership role on AI,” citing ties to the EU, U.S., and China plus expertise including the AI Safety Institute, so it could act as an “honest broker,” a middle power claiming it can talk to all sides without being locked to one camp. Burnham said: “I think we can really do something in this space [that is] very significant.”

Burnham was due to meet Donald Trump for the first time as prime minister on Tuesday. He said he will assure Trump the UK will take a balanced approach: clear-eyed about benefits and where to be careful. Politico notes Trump has warned of AI risks and has also branded existential warnings a “HOAX.”

Burnham said insider warnings have “made me pause,” but he wants a “sweet spot” and not to become a “doom-monger about AI.” UK AI minister Kanishka Narayan traveled with him. No. 10 said AI would dominate diplomatic engagements in New York.

Intent for a G20 deal.

[Aaron Harme @Aaron_Harme]

At the UN General Assembly, Trump decided the branding problem with artificial intelligence is the word artificial. Official U.S. documents, he said Tuesday morning, should say “super intelligence” instead.

The Verge reports Richard Lawler on the Tuesday UNGA speech: Trump claimed the United States is officially renaming artificial intelligence, arguing the word “artificial” makes intelligence fake. Speech color reframes “whoever wins AI” as “whoever wins…SI / super intelligence,” and “The United States leads the world in super intelligence…”

Verge quotes Trump on the documents pitch: “From this point forward, all of the United States documents, and hopefully the world’s, will be changed to use the much more accurate term, super as opposed to artificial.”

Futurism quotes Trump: “From this point forward, all of United States documents, and hopefully the world’s will be changed to use the much more accurate term ‘super’ as opposed to artificial… So it’s super intelligence.” Futurism also notes a Saturday Truth Social post floating Superior Intelligence (SI), Extreme Intelligence (EI), or Supreme Intelligence (SI) before the UN speech made “super intelligence” the documents pitch.

A branding order claimed from the podium.

[Aaron Harme @Aaron_Harme]

@minchoi How do you have Grok Bot communicating with Claude and Codex?