[AI WEEKLY RUNDOWN] White House Blocks GPT-5.6, China Bypasses Chip Bans, and Altman Demands $1 Trillion (June 21 – June 28, 2026)

DjamgaMind: Audio Intelligence for the C-Suite (Daily AI News, Energy, Healthcare, Finance)

Full-Stack AI Intelligence. Zero Noise.The definitive audio briefing for the C-Suite and AI Architects. From Daily News and Strategic Deep Dives to high-density Industrial & Regulatory Intelligence—decoded at the speed of the AI era. . 👉 Start your specialized audio briefing today at Djamgamind.com

AI Jobs and Career

I wanted to share an exciting opportunity for those of you looking to advance your careers in the AI space. You know how rapidly the landscape is evolving, and finding the right fit can be a challenge. That's why I'm excited about Mercor – they're a platform specifically designed to connect top-tier AI talent with leading companies. Whether you're a data scientist, machine learning engineer, or something else entirely, Mercor can help you find your next big role. If you're ready to take the next step in your AI career, check them out through my referral link: https://work.mercor.com/?referralCode=82d5f4e3-e1a3-4064-963f-c197bb2c8db1. It's a fantastic resource, and I encourage you to explore the opportunities they have available.

• Full Stack Engineer [$150K-$220K]
• Software Engineer, Tooling & AI Workflow, Contract [$90/hour]
• DevOps Engineer, India, Contract [$90/hour]
• More AI Jobs Opportunitieshere

Job Title	Status	Pay
Full-Stack Engineer	Strong match, Full-time	$150K - $220K / year
Developer Experience and Productivity Engineer	Pre-qualified, Full-time	$160K - $300K / year
Software Engineer - Tooling & AI Workflows (Contract)	Contract	$90 / hour
DevOps Engineer (India)	Full-time	$20K - $50K / year
Senior Full-Stack Engineer	Full-time	$2.8K - $4K / week
Enterprise IT & Cloud Domain Expert - India	Contract	$20 - $30 / hour
Senior Software Engineer	Contract	$100 - $200 / hour
Senior Software Engineer	Pre-qualified, Full-time	$150K - $300K / year
Senior Full-Stack Engineer: Latin America	Full-time	$1.6K - $2.1K / week
Software Engineering Expert	Contract	$50 - $150 / hour
Generalist Video Annotators	Contract	$45 / hour
Generalist Writing Expert	Contract	$45 / hour
Editors, Fact Checkers, & Data Quality Reviewers	Contract	$50 - $60 / hour
Multilingual Expert	Contract	$54 / hour
Mathematics Expert (PhD)	Contract	$60 - $80 / hour
Software Engineer - India	Contract	$20 - $45 / hour
Physics Expert (PhD)	Contract	$60 - $80 / hour
Finance Expert	Contract	$150 / hour
Designers	Contract	$50 - $70 / hour
Chemistry Expert (PhD)	Contract	$60 - $80 / hour

🎧 Listen ADS-FREE: https://podcasts.apple.com/us/channel/djamgamind/id6760446113

Visit our Research Hub at https://djamgamind.com/pdfs

#DJAMGAMIND

Summary: In this weekly briefing, we analyze “The Sovereignty Veto and the Geopolitical Tech Rift.” We deconstruct the White House’s unprecedented control over frontier AI releases, forcing OpenAI’s GPT-5.6 into a restricted preview and dictating the partial return of Anthropic’s Mythos 5. We dissect the escalating trade and component war, tracking China’s launch of the GPU-free LineShine supercomputer, Beijing’s ban on 46 US defense firms, and President Trump’s 100% tariff threat over European digital taxes. We explore Sam Altman delaying OpenAI’s IPO over a strict $1 trillion valuation demand, SpaceX’s $6.3 billion compute deal with Reflection AI, and Samsung’s staggering $648 billion domestic investment plan.

Important Topics:

• White House Gating GPT-5.6 Release: OpenAI restricts the launch of its GPT-5.6 series (Sol, Terra, Luna) to a small group of approved partners following direct intervention by the Office of the National Cyber Director.

• Anthropic Mythos 5 Partially Revived: The Commerce Department approves Anthropic to resume selling Mythos 5 to a restricted list of over 100 vetted cyber defenders, leaving the consumer Fable 5 model banned.

• China Reclaims Supercomputer Crown: The National Supercomputer Center in China bypasses US chip embargoes to build the LineShine machine, hitting 2.198 Exaflops using a custom GPU-free, 304-core processor architecture.

• Sam Altman Demands $1 Trillion IPO Valuation: OpenAI advisors delay the company’s planned IPO to 2027 after CEO Sam Altman rejects any valuation below $1 trillion as a “nonstarter,” causing SoftBank shares to plunge 13%.

• Trump Threatens 100% Tariff on Tech Taxes: President Trump warns of a 100% tariff on goods from any country attempting to levy digital services taxes against American tech monopolies like Meta and Alphabet.

  

• China Retaliates Against US Tech Bans: Beijing bans 46 US firms, mostly defense contractors, and places rare earth miners on export control lists in a tit-for-tat response to the Pentagon’s Section 1260H military blacklist.

• SpaceX Secures $6.3B Compute Deal: Reflection AI signs a multi-year, $150-million-a-month contract to rent Nvidia GB300 compute infrastructure from SpaceX’s Colossus data centers.

• Apple Lobbies to Buy Blacklisted Chinese Chips: Apple asks the Trump administration for permission to purchase memory chips from the blacklisted Chinese manufacturer CXMT to offset surging AI-driven component costs.

• Samsung Plans $648 Billion Domestic Investment: Local media reports indicate Samsung will commit 1,000 trillion won over the next decade to build chip factories, AI data centers, and battery plants across South Korea.

• Five Eyes Issue Severe AI Cyber Warning: Global intelligence agencies warn that AI models capable of dismantling national and corporate cyber defenses are “months, not years” away, demanding immediate zero-trust compliance.

• IBM Claims Sub-1 Nanometer Chip Breakthrough: IBM unveils a 0.7 nm “nanostack” architecture, doubling transistor density to deliver a 70% boost in energy efficiency for advanced AI execution.

• TikTok Transitions into a Fintech Super App: TikTok applies for financial licenses in Brazil to offer credit and prepaid accounts, expanding its platform beyond video into a dominant transactional super app.

🔗 RESOURCES

• AI Learning App Recommendation: AI & ML Tutor PRO https://apps.apple.com/ca/app/ai-ml-tutor-pro/id1610947211

• DJAMGATECH: Carrer Booster – Master AWS, Azure, AI & GCP Certifications | https://apps.apple.com/ca/app/djamgatech-ai-cert-exams-prep/id1560083470

• Read Aloud wonderland bedtime Adventures:

https://wonderland-bedtime-stories.web.app/

⚗️ PRODUCTION NOTE: We Practice What We Preach.

AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow.

OpenAI unveils GPT-5.6 in limited preview LINK

• OpenAI has begun previewing its GPT‑5.6 series to a small group of trusted partners at the government’s request, but says this kind of government access process should not become the long‑term default.

• The preview follows Trump’s AI cybersecurity order this month, which asks companies to submit their most powerful models for voluntary government review 30 days before any public release, with Meta the only holdout.

• The series has three versions: Sol, the strongest, with a new “max” reasoning effort, costs $5 per million input tokens and $30 for output; Terra and Luna are cheaper everyday and low‑cost options.

  Smartphone 101 - Pick a smartphone for me - android or iOS - Apple iPhone or Samsung Galaxy or Huawei or Xaomi or Google Pixel
  

White House lets Anthropic revive Mythos 5 LINK

• The U.S. Commerce Department approved Anthropic’s request on June 26 to resume selling Claude Mythos 5, partially reversing an export-control order that had forced the company to pull its strongest models offline two weeks earlier.

• Mythos 5 can now go to a small group of cyber defenders and infrastructure providers, with more than 100 approved companies and institutions, including Fortune 500 firms, plus non-U.S. employees of those registered partners.

• The letter from Commerce Secretary Howard Lutnick said Anthropic made “significant progress” reducing risks, but it did not mention Fable 5, the consumer version, which stays offline with no set date for its return.

Apple asks Trump to clear blacklisted Chinese chips LINK

• Apple has reportedly asked the Trump administration to approve buying memory chips from Chinese maker ChangXin Memory Technologies, or CXMT, even though the company sits on a Pentagon blacklist, the Financial Times reports.

• Apple first approached the Commerce Department more than a month ago and has since lobbied other officials and allies in Washington, arguing that sharply rising memory chip costs are straining the company’s finances.

• CXMT was named a Chinese military company by the Defence Department under Biden and added to the Commerce Department’s Entity List, which bars US firms from exporting to it without a licence that is usually denied.

TikTok is quietly becoming a super app LINK

• TikTok is steadily turning into a “super app,” a single platform where people can do far more than watch videos, after adding TikTok Shop, hotel booking, search, games, and a push for a fintech license.

• In May, TikTok launched TikTok GO in the United States, letting users discover and book hotels, attractions, and experiences through videos, search, and location pages, which puts it in direct competition with Google Search and Google Maps.

• The company applied to Brazil’s central bank for two licenses, one to offer prepaid accounts for storing funds and making payments, and another to act as a direct credit provider that lends money or connects borrowers and lenders.

Trump threatens 100% tariff over tech taxes LINK

• President Trump threatened on Friday to place a “100% TARIFF” on goods from any country that charges a digital services tax on American companies, posting the warning on Truth Social.

• Trump said this tariff would “supersede Trade Deals made with the Country, whether implemented, signed, or not,” and his post singled out “Numerous European Countries” he says are considering the taxes.

• The digital services taxes usually apply only to the largest tech companies like Meta, Alphabet and Amazon, all U.S. firms, and more than a dozen countries have already imposed them.

  
  
  

Altman won’t go public for less than $1 trillion LINK

• OpenAI is leaning toward pushing its IPO to 2027 because CEO Sam Altman insists on a $1 trillion valuation, up from the company’s last private figure of $730 billion, which advisors say is the main reason for the delay.

• Advisors gave Altman two choices: go public in 2027 at the trillion-dollar price, or move faster at a lower valuation, but he rejected anything below a trillion as a “nonstarter,” even as choppy tech markets worry investors.

• The delay hit SoftBank hard, sending its stock down 13 percent, its steepest drop since August 2024, since the Japanese investor’s stake in OpenAI is set to reach about $65 billion by October.

SpaceX reportedly plans a Starlink mobile service LINK

• SpaceX is reportedly planning a Starlink-branded retail mobile service for US consumers, a move that would put the company in direct competition with AT&T, Verizon, and T-Mobile.

• SpaceX President Gwynne Shotwell told investors during an IPO roadshow that the company is considering the service, and could even build its own wireless network on the ground to support it.

• The plan follows the FCC’s approval last month of SpaceX’s $17 billion purchase of 65 MHz of wireless spectrum from EchoStar, plus earlier trademark filings for “Starlink Mobile” and “Powered by Starlink.”

White House asks OpenAI to delay GPT-5.6 LINK

• OpenAI plans to release its new model, GPT 5.6, only to a small group of close partners rather than the public, after the Trump administration told the company to limit access, according to The Information.

• CEO Sam Altman reportedly told staff the government would be “approving access customer by customer” during a preview period, and OpenAI hopes to follow with a broader release a “couple of weeks later” if it goes well.

• The Office of the National Cyber Director and the Office of Science and Technology Policy reportedly asked for the limited release, and OpenAI staffers worked closely with the government on the upcoming launch.

Apple’s most powerful Macs may wait until 2027 LINK

• Apple’s most powerful Macs could be delayed until 2027, as the company reportedly plans to skip the M6 Pro and Max chips and move straight to the M7 chip, according to Bloomberg’s Mark Gurman.

• The M7 chips are expected to launch in the first half of 2027 with a focus on on-device AI processing, marking the first time Apple offers only a base version without Pro or Max variations.

• Apple may release the M7 Max and Pro chips by the end of 2027, while an M7 Ultra chip could arrive in 2028, since the company has not put out an Ultra model since the M3 generation.

New US bill would force AI risk reporting LINK

• A new US bill called the AI Incident Reporting Act would require developers of high-risk “covered models” to report major safety and security incidents to the Commerce Department, creating a federal oversight system for advanced AI.

• Developers would have to file an initial report within seven days of learning of an incident, and for imminent threats, the Commerce Department must alert congressional leaders within 48 hours of getting it.

• Reportable incidents include models that evade human oversight, resist shutdown, or enable cyberattacks and chemical weapons, while violators face civil penalties of up to $2 million, with each day counted as a separate violation.

Samsung plans $648 billion South Korea bet LINK

• Samsung is expected to announce on June 29 a plan to invest 1,000 trillion won, about $647.5bn, across South Korea over the next 10 years, one of the country’s largest corporate investment commitments.

• Of the total, a reported 300 trillion won would build chip factories in the country’s southwest, a region left out of the semiconductor boom, while the rest covers AI data centres, batteries, and displays.

• The pledge comes from local media reports, not a formal Samsung statement, and rival SK Hynix is expected to unveil its own investment plans at the same meeting with President Lee Jae Myung.

Apple hikes MacBook and iPad prices, warns of more LINK

• Apple raised prices on multiple MacBook and iPad models after saying surging memory and storage costs have made it impossible to keep prices unchanged.

• The company blamed the AI boom for creating unprecedented demand for memory chips, causing component prices to soar and forcing Apple to pass some of those higher costs on to consumers.

• Apple’s stock fell about 5% following the announcement, and the company warned that additional price increases across other products could happen if component costs remain elevated.

Anthropic accuses Alibaba of biggest-ever Claude theft LINK

• Anthropic accused Alibaba of carrying out its largest known AI “distillation” attack, alleging the company used around 25,000 fraudulent accounts to generate 28.8 million interactions with its AI models.

• The AI startup says the campaign was an attempt to illegally extract its technology and has urged the U.S. government and the AI industry to work together to better protect American AI capabilities.

• The allegations come as AI security and export controls become a bigger focus, with Anthropic also facing recent U.S. government restrictions on access to some of its latest AI models.

IBM claims world’s first sub-1 nanometer chip LINK

• IBM unveiled the world’s first sub-1 nanometer chip technology, featuring a new 0.7 nm “nanostack” architecture that nearly doubles transistor density compared to its previous 2 nm chip.

• The new design is expected to deliver up to 50% better performance or 70% greater energy efficiency, making it well suited for AI, cloud computing, and next-generation electronic devices.

• IBM believes the breakthrough can extend chip scaling for at least another decade, with commercial production of the technology potentially beginning within the next five years.

Tesla teams up to build largest US power plant LINK

• Sunrun, Renew Home, and Tesla announced a partnership to combine millions of home batteries, smart thermostats, and EV systems into a virtual power plant capable of delivering more than 16 gigawatts of flexible electricity.

• The companies say the system can help power AI data centers and other large electricity users within months, reducing strain on the grid without requiring major new infrastructure.

• By using existing home energy devices more efficiently, the partnership aims to lower electricity costs for households, improve grid reliability, and accelerate the rollout of new AI and data center capacity.

AI researchers continue to leave Google for its rivals LINK

• Two more researchers who helped build Google’s Gemini model, Jonas Adler and Alexander Pritzel, are leaving the company for Anthropic, according to a Bloomberg report on their departures.

• These exits add to a worrying pattern at Google, which last week saw researcher Noam Shazeer leave for OpenAI and DeepMind Director John Jumper announce his own move to Anthropic.

• The article suggests the trend may keep going as OpenAI and Anthropic prepare to go public, giving both companies a strong chance to recruit top AI talent by offering equity.

OpenAI reveals its first AI processor LINK

• OpenAI has revealed Jalapeño, its first custom chip built in partnership with Broadcom, an “intelligence processor” designed to run AI inference for current and future large language models on AI servers.

• Jalapeño is an ASIC made for AI inference, the process where models handle a user’s request to run an agent like Codex or give a ChatGPT response, rather than training on data.

• OpenAI calls the chip the first step in a multi-generation compute platform that it plans to deploy by the end of 2026, with early testing showing better performance per watt than current options.

Anthropic puts Claude inside your Slack channels LINK

• Anthropic has launched Claude Tag, a product that puts Claude directly inside Slack channels as a permanent team member that gathers knowledge over time, works on its own schedule, and can act without being tagged first.

• Available in beta for Enterprise and Team plans, Claude Tag gives each channel one shared Claude with its own identity and accounts, so any team member can steer, correct, and build on a task in real time.

• Administrators define what Claude can access at the workspace level through “agent identity,” set token spending ceilings per channel, and keep shared channels from becoming a backdoor into anyone’s private documents since Claude uses its own accounts.

Meta is building a prediction markets app LINK

• Mark Zuckerberg has told a small Meta team to build a smartphone app called Arena that would compete with prediction market platforms Polymarket and Kalshi, operating separately from Facebook and Instagram, the New York Times reported Tuesday.

• Arena would start with a video-game-style points system instead of real-money wagers, which sidesteps oversight from the Commodity Futures Trading Commission, though Meta has not ruled out adding real-money betting later.

• Meta could steer its roughly 3 billion monthly active users toward Arena, a scale that dwarfs existing platforms, and the news sent DraftKings down more than 2 percent while Flutter Entertainment and Robinhood also fell.

SpaceX secretly launches its new Starfall capsule LINK

• SpaceX quietly flew its new Starfall return capsule for the first time on June 23, launching aboard a Falcon 9 from Space Launch Complex 40 at Cape Canaveral and cutting the webcast about 10 minutes after liftoff.

• According to an FAA assessment, the disk-shaped, uncrewed Starfall stands 2.5 feet tall, weighs roughly 4,600 pounds, carries 2,200 pounds of payload, and uses a detachable carbon-fiber heat shield filled with compressed nitrogen gas.

• The capsule carries no propulsion and cannot deorbit itself, and SpaceX is targeting a Pacific Ocean splashdown about 700 nautical miles off the West Coast, though it has not said how long the vehicle will stay in orbit.

Alibaba sues Pentagon to escape military blacklist LINK

• Alibaba has filed a lawsuit against the US Department of Defense to be taken off the Pentagon’s 1260H list, which names companies allegedly tied to the Chinese People’s Liberation Army.

• The Defense Department labeled Alibaba a “military-civil fusion contributor to the Chinese defense industrial base” over its ties to Beijing regulators, while the updated list also includes internet services provider Baidu.

• Being on the list bars the Defense Department from doing business with Alibaba through third parties, and the company says it presented evidence it does not back the Chinese military but got no response.

Trump signs order to boost quantum research LINK

• President Trump signed two executive orders on Monday meant to push research into quantum computing forward and tighten security practices in preparation for the risks a post-quantum world could bring to online systems.

• One order calls for building a quantum computer strong enough for scientific research, tells federal agencies to plan for quantum-enabled technology, and seeks to boost domestic supply chains for quantum infrastructure and materials.

• The second order aims to speed migration to post-quantum cryptography by 2031, protecting systems against quantum threats that could break the encryption algorithms long used to secure private information online.

Meta launches its own smart glasses LINK

• Meta is launching its own line of smart glasses called Meta Glasses, dropping the Ray-Ban and Oakley branding entirely while still making them with EssilorLuxottica and selling them through the same retail stores.

• The new glasses start at $299, undercutting the $379 Ray-Ban Gen 2 and $499 Scriber and Blazer Optics models, while keeping the same comfort fit, battery life, camera quality, and dual-camera AI rocker button.

• The lineup includes the chunky “Fury” frame, the more compact “Adventurer,” and a Kylie Jenner design called the Starfire Kylie Edition, which has oval lenses, an embedded gem, and a charging case with a vanity mirror.

OpenAI’s ‘Patch the Planet’ takes on Anthropic LINK

• OpenAI has teamed up with cybersecurity firm Trail of Bits on a program called Patch the Planet, which uses AI to find and fix security flaws in widely used open-source software that enterprises depend on.

• The program starts each engagement by consulting maintainers, then uses OpenAI’s models and Codex Security to investigate vulnerabilities, while Trail of Bits engineers review findings to filter out false positives before they reach the projects.

• Initial participants include Python, Go, cURL, and python.org, and the work has already identified hundreds of security issues and merged dozens of patches, with HackerOne and Calif supporting triage and disclosure.

China retakes world’s fastest supercomputer crown LINK

• China has taken back the title of world’s fastest supercomputer for the first time since 2017, with its LineShine machine hitting 2.198 Exaflops and beating the previous leader, El Capitan, at the Lawrence Livermore National Laboratory.

• LineShine got around US technology embargoes by avoiding GPUs entirely, instead running on a custom 304-core processor with 13.79 million cores at 1.55GHz, drawing 42.2 megawatts for an efficiency of 52.07 Gigaflops per watt.

• The previously unlisted machine, built by the National Supercomputer Center without public funding, kept some details hidden, including which company made the CPUs and the type of chip technology used in its design.

Instagram takes on Netflix with TV app LINK

• Instagram is going after streaming services like Netflix and Amazon Prime Video by adding longer videos, episodic series, and Live TV to the TV app it launched last year, aiming to reach viewers in the living room.

• The TV app is rolling out to Samsung TVs, joining Amazon Fire and Google TVs, and is gaining channels tailored to creators and topics users like, plus the ability to cast Reels and Saved content from a phone.

• A dedicated section for horizontal videos is being tested to make watching on TV easier, and people can now view other users’ Stories within the app, which previously only played Reels natively.

SpaceX lands $6.3B compute deal with Reflection AI LINK

• SpaceX has signed a $6.3 billion deal to sell computing power to Reflection AI, an open-source artificial intelligence startup that becomes the latest outside customer to tap Elon Musk’s Colossus infrastructure.

• Reflection gets immediate access to Nvidia GB300s, the chips used to train and run AI models, and agreed to pay SpaceX $150 million per month from July 1, 2026 through 2029, with either side able to exit on 90 days’ notice.

• The deal adds to SpaceX’s computing power agreements with Anthropic, Google and Cursor, which Musk’s company is now acquiring, and brings in an open-source lab as enterprises reassess relying on closed AI systems.

China imposes restrictions on US rare earth companies LINK

• China placed 10 U.S. industrial suppliers on its export control list and barred 46 American firms, mostly defense contractors, from government procurement, retaliating against the Pentagon adding Chinese companies to its 1260H military blacklist.

• The export control list targets rare earth miners MP Materials and USA Rare Earth, drone makers Teal Drones and Jaia Robotics, plus Oshkosh Defense and Ball Aerospace, blocking exports of any dual-use items originating in China to them.

• Analysts called the countermeasures largely symbolic since most targeted firms have little business in China, while the Pentagon’s 1260H list bars the Defense Department from awarding direct contracts to listed companies starting June 30.

WhatsApp gets a new chief LINK

• Meta is putting $900 million into the Indian fintech startup Cred and plans to make its founder, Kunal Shah, the new head of WhatsApp, replacing Will Cathcart after roughly seven years.

• The deal gives Meta about a 20% stake in Cred, which runs an app that rewards customers for paying credit card bills on time, valuing the company at $4.5 billion post-money.

• Shah, recruited by product chief Chris Cox and relocating from Bangalore to Menlo Park, will lead efforts on WhatsApp’s advertising, subscriptions, and the work of adding AI agents to the messaging app.

Microsoft may host China’s DeepSeek LINK

• Microsoft’s CEO Satya Nadella said the company is weighing whether to host a version of DeepSeek, the ultra-low-cost Chinese model that OpenAI and Anthropic accuse of copying their work.

• Hosting DeepSeek would send much more traffic to the Chinese model-maker and add pressure on OpenAI and Anthropic, two rivals already caught in a long price war over their proprietary models.

• The move fits Microsoft’s recent push, which included rolling out a suite of low-cost models aimed at customers hit by soaring AI bills, while Nadella argued value should not flow to only a few models.

New report warns AI could topple governments LINK

• The Five Eyes intelligence agencies from Australia, the US, the UK, New Zealand, and Canada issued a rare joint statement warning that AI models powerful enough to take down governments and businesses are only months away, urging leaders to “act now”.

• The statement said frontier AI models would accelerate the speed, scale, and sophistication of cyber threats, lowering barriers for bad actors, and called for a whole-of-society response treating cyber risk as a core business and leadership responsibility.

• No companies were named, but attention centres on Anthropic’s models Mythos and Fable 5, both suspended for use by “foreign nationals” in June by the US government, which cited advice from national security authorities.

Getty Images soars 120% on OpenAI deal LINK

• Getty Images saw its shares climb roughly 120 percent after announcing a multi-year deal with OpenAI to display its licensed photo libraries inside ChatGPT’s search and discovery features.

• Under the agreement, when a user asks ChatGPT a question that calls for an image, the chatbot can return a licensed Getty photograph rather than a synthetic one generated by AI.

• The companies disclosed no financial terms, no revenue split, and nothing about whether Getty’s images may train future OpenAI models, with the deal limited to display rather than training.

OpenAI holds back GPT-5.6 series:

The AI giant announced a “limited preview” of its new GPT-5.6 family of frontier AI models. That includes Sol, the flagship; Terra, a comparable model to GPT-5.5 that eats up only half the tokens; and Luna, an even lower-cost option. For now, only a few trusted partners have access to these models, as requested/instructed by the US government. The models will apparently open up access to a larger population within the coming weeks. In their post, OpenAI leadership pushes back gently, suggesting that direct government intervention should not become “the long-term default” for new AI model releases, as it “keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them.” Former White House AI policy adviser Dean W. Ball was harsher in his criticism, positing that “you should not expect the most powerful people in the world using the most powerful technology ever conceived in a way that is inscrutable to the public to turn out well, and you should see that dynamic as fundamentally inconsistent with a democratic republic.”

About that OAI IPO…

The New York Times reports that OpenAI leadership is having second thoughts about its IPO plans, at least for the remainder of 2026. According to a trio of unnamed sources, CEO Sam Altman wants the value to exceed $1 trillion as a public company, and it’s unclear if that’s currently possible amid unpredictable global markets, lingering concerns about future AI revenues, and of course the downward slide of SpaceX stock following its own blockbuster debut. No word yet on whether or not Anthropic, the Coca-Cola to OpenAI’s Pepsi (or is it the other way around?), are also having IPO cold feet.