Government gives greenlight for new AI models
The White Home asserts new management over superior AI fashions, with tech giants like OpenAI and Anthropic agreeing to governmental oversight. Fox Enterprise Correspondent Connor Hansen experiences on the administration’s coverage shift. OpenAI CEO Sam Altman voices issues in regards to the authorities choosing prospects for AI instruments, emphasizing nationwide safety.
OpenAI’s newest AI model is getting a gatekeeper earlier than most individuals can attempt it: the U.S. authorities.
The corporate says GPT-5.6 Sol, its latest flagship mannequin, will begin with a restricted preview for a small group of trusted companions. OpenAI stated it previewed the mannequin’s capabilities with the U.S. authorities forward of launch and, on the authorities’s request, is beginning with a restricted group of trusted companions whose participation has been shared with the federal government earlier than a broader rollout. The mannequin will reportedly be accessible solely to prospects authorised by the Trump administration throughout this preliminary preview interval.
MOST PROMINENT AI CHATBOTS HAVE LIBERAL BIAS, NEW STUDY FINDS
That to me is an enormous shift. Normally, you hear a few new AI mannequin and marvel when you should use it. Now, the larger query could also be who will get entry first and the way these selections are made.
Join my FREE CyberGuy Report
- Get my greatest tech suggestions, pressing safety alerts and unique offers delivered straight to your inbox.
- For easy, real-world methods to identify scams early and keep protected, go to CyberGuy.com – trusted by thousands and thousands who watch CyberGuy on TV every day.
- Plus, you may get instantaneous entry to my Final Rip-off Survival Information free once you be a part of.
OpenAI is limiting entry to GPT-5.6 Sol as U.S. officers evaluation potential cybersecurity dangers tied to the highly effective new AI mannequin. (Kurt “CyberGuy” Knutsson)
What’s OpenAI’s GPT-5.6 Sol?
GPT-5.6 Sol is a part of OpenAI’s new GPT-5.6 lineup, which additionally contains Terra and Luna. OpenAI describes Sol as its strongest mannequin but, with enhancements in coding, biology and cybersecurity duties. In its public announcement, OpenAI stated GPT-5.6 fashions will initially be accessible by means of the API and Codex to a choose group of trusted companions and organizations, with broader availability deliberate for ChatGPT, Codex and the API quickly. As of early July, OpenAI says GPT-5.6 isn’t accessible in ChatGPT throughout the preview, and the corporate has not introduced a general-availability date.
The cybersecurity part is what has Washington paying shut consideration. OpenAI says Sol is best at serving to individuals discover and repair software program vulnerabilities than reliably finishing up full cyberattacks. The corporate additionally says Sol doesn’t cross its inside “Cyber Vital” threshold, though it admits benchmark exams can not predict each doable use when the mannequin is mixed with different instruments.
That’s the concern. The identical AI that may assist a safety crew repair a software program flaw might additionally give scammers and hackers a head begin if the guardrails fail.
Why the federal government is stepping in
The Trump administration has been constructing a brand new AI cybersecurity framework. A June government order says superior AI can strengthen the nation, however it additionally creates national security concerns that require coordination throughout federal companies and personal corporations.
The White Home reality sheet says the order creates a voluntary framework for lined frontier fashions. It additionally says the federal government could get safe early entry for trusted companions, whereas stating the order doesn’t authorize obligatory licensing, pre-clearance or permits for releasing AI fashions.
That distinction is vital. The administration says that is about safety. Critics fear it might flip into quiet authorities management over who will get the latest AI instruments first.
Anthropic was pulled into the identical AI struggle
OpenAI isn’t the one AI firm that has confronted authorities strain over highly effective AI fashions. On June 12, Anthropic stated the U.S. authorities directed it to droop entry to its Fable 5 and Mythos 5 fashions by international nationals. Anthropic stated the order pressured it to disable the fashions extra broadly so it might comply.
CyberGuy beforehand lined Anthropic’s Mythos AIand why its potential to seek out software program vulnerabilities raised huge cybersecurity questions. That helps clarify why these fashions turned a part of the bigger debate over AI, cyber danger and who will get entry first.
Since then, the standoff has modified. The U.S. Commerce Division later lifted export controls on Anthropic’s Fable 5 and Mythos 5 fashions. Anthropic stated it could restore entry, and Fable 5 was reportedly again on-line as of July 1. Stories say Anthropic agreed to strengthen safeguards, work with the U.S. authorities on model-release protocols and report malicious exercise tied to the fashions. Reuters additionally reported that the restrictions had been eliminated after talks between Anthropic and U.S. officers.
That makes the Anthropic episode a helpful comparability for OpenAI’s restricted rollout. In each instances, the problem is now not solely about what an AI mannequin can do. Additionally it is about how rapidly highly effective AI instruments needs to be launched, who ought to get early entry and what position the federal government ought to play.
The Trump administration’s AI cybersecurity push is elevating new questions on who will get early entry to superior synthetic intelligence. (Samuel Boivin/NurPhoto through Getty Photos)
Why this reaches past Silicon Valley
Right here is the larger query: how ought to highly effective AI instruments be launched after they will help shield hospitals, banks and companies, however might additionally assist the mistaken individual discover safety holes sooner?
Safety groups need higher instruments as a result of they’re already stretched skinny. Hospitals and health care systems,banks, local utilities and critical systems, plus businesses all run on software program that may have hidden holes. A stronger AI mannequin might assist discover these weak spots earlier than criminals exploit them.
Nevertheless, the entry query is sophisticated. Who will get early entry? Who will get informed to attend? What occurs if a small safety firm wants the identical AI instrument as a serious authorities contractor?
For now, the talk comes right down to balancing safety, entry and accountability as these AI instruments transfer nearer to the providers you depend on day-after-day.
How you can shield your self as AI will get extra highly effective
AI instruments are getting extra succesful, and meaning scammers will preserve utilizing the thrill round them to trick you. Listed here are the steps to take now.
1) Be careful for faux early-access provides
At any time when a serious AI product will get consideration, imposters observe. Watch out with emails, texts or advertisements that promise early entry to GPT-5.6 Sol, Mythos 5 or any restricted AI mannequin. Go on to the official firm web site as a substitute of clicking a hyperlink.
2) Use sturdy antivirus software program
Sturdy antivirus software program will help shield you from malicious hyperlinks, phishing websites and downloads that scammers could disguise as AI instruments or beta invites. The easiest way to guard your self from malicious hyperlinks that set up malware, doubtlessly accessing your non-public info, is to have sturdy antivirus software program put in on all of your units. Get my picks for the very best 2026 antivirus safety winners to your Home windows, Mac, Android & iOS units at Cyberguy.com
3) Use a password supervisor
Use sturdy, distinctive passwords to your vital accounts. A password manager will help you create and retailer advanced passwords, so you aren’t reusing the identical login throughout a number of websites.
4) Activate two-factor authentication
Activate two-factor authentication (2FA) every time it’s accessible. Authentication apps or safety keys are stronger than textual content message codes as a result of they’re more durable for scammers to intercept.
OPENAI TIGHTENS AI RULES FOR TEENS BUT CONCERNS REMAIN
OpenAI says GPT-5.6 Sol will help discover and repair software program flaws, however officers fear highly effective AI instruments might additionally assist hackers. (Picture by Philip Dulian/image alliance through Getty Photos)
5) Hold your units and apps up to date
Hold your telephone, pc, browser and apps up to date. Many cyberattacks goal recognized software program flaws that have already got patches, so updates are one of many best methods to shut safety gaps.
6) Watch out what you add into AI instruments
Assume twice earlier than pasting delicate info into any AI instrument. Keep away from importing Social Safety numbers, monetary data, medical paperwork or non-public firm information except precisely how the service handles your information.
7) Take away your private information from people-search websites
Scammers could make AI-themed attacks extra convincing after they can discover your house handle, telephone quantity, kin or office on-line. A knowledge removing service will help cut back how a lot of your private info seems on information dealer and people-search websites. Try my prime picks for information removing providers and get a free scan to seek out out in case your private info is already out on the internet by visiting Cyberguy.com
8) Decelerate earlier than clicking pressing AI invitations
Scammers love urgency. If a proposal claims you might want to act quick to get entry to a brand new AI mannequin, deal with that as a warning signal. Seek for the announcement your self and confirm it from the corporate’s official web site.
Watch the CyberGuy Dwell replay: Lock Down Your Telephone in 30 Minutes
Your telephone holds your e mail, passwords, photographs, banking apps and private information. On this free CyberGuy Dwell replay, Kurt the CyberGuy walks you step-by-step by means of easy telephone safety fixes you are able to do at your personal tempo. You’ll learn to enhance your privateness settings, spot the most recent telephone scams, use trusted safety instruments and stroll away with a easy guidelines to remain protected. Watch the replay and get our guidelines right here: CyberGuyLive.com
Kurt’s key takeaways
AI has reached a degree the place Washington desires a better look earlier than everybody will get entry. I perceive why cybersecurity officers are paying consideration. A mannequin that may spot software program flaws might assist shield a hospital, a financial institution or a utility. Within the mistaken arms, although, that very same functionality might create new dangers. For you, the takeaway is simple. AI instruments will preserve getting extra succesful, and that may be excellent news after they assist defend your information. Nevertheless, you continue to want to guard your accounts, query shock AI provides and assume twice earlier than sharing delicate info with any new instrument.
Ought to highly effective AI instruments face a safety evaluation earlier than public launch, or ought to corporations resolve when they’re prepared for everybody to make use of? Tell us by writing to us at Cyberguy.com
CLICK HERE TO DOWNLOAD THE FOX NEWS APP
Join my FREE CyberGuy Report
- Get my greatest tech suggestions, pressing safety alerts and unique offers delivered straight to your inbox.
- For easy, real-world methods to identify scams early and keep protected, go to CyberGuy.com – trusted by thousands and thousands who watch CyberGuy on TV every day.
- Plus, you may get instantaneous entry to my Final Rip-off Survival Information free once you be a part of.
Copyright 2026 CyberGuy.com. All rights reserved.
