Technology
OpenAI admits AI browsers face unsolvable prompt attacks
NEWYou can now listen to Fox News articles!
Cybercriminals don’t always need malware or exploits to break into systems anymore. Sometimes, they just need the right words in the right place. OpenAI is now openly acknowledging that reality. The company says prompt injection attacks against artificial intelligence (AI)-powered browsers are not a bug that can be fully patched, but a long-term risk that comes with letting AI agents roam the open web. This raises uncomfortable questions about how safe these tools really are, especially as they gain more autonomy and access to your data.
Sign up for my FREE CyberGuy Report
Get my best tech tips, urgent security alerts, and exclusive deals delivered straight to your inbox. Plus, you’ll get instant access to my Ultimate Scam Survival Guide — free when you join my CYBERGUY.COM newsletter.
NEW MALWARE CAN READ YOUR CHATS AND STEAL YOUR MONEY
AI-powered browsers can read and act on web content, which also makes them vulnerable to hidden instructions attackers can slip into pages or documents. (Kurt “CyberGuy” Knutsson)
Why prompt injection isn’t going away
In a recent blog post, OpenAI admitted that prompt injection attacks are unlikely to ever be completely eliminated. Prompt injection works by hiding instructions inside web pages, documents or emails in ways that humans don’t notice, but AI agents do. Once the AI reads that content, it can be tricked into following malicious instructions.
OpenAI compared this problem to scams and social engineering. You can reduce them, but you can’t make them disappear. The company also acknowledged that “agent mode” in its ChatGPT Atlas browser increases risk because it expands the attack surface. The more an AI can do on your behalf, the more damage it can cause when something goes wrong.
OpenAI launched the ChatGPT Atlas browser in October, and security researchers immediately started testing its limits. Within hours, demos appeared showing that a few carefully placed words inside a Google Doc could influence how the browser behaved. That same day, Brave published its own warning, explaining that indirect prompt injection is a structural problem for AI-powered browsers, including tools like Perplexity’s Comet.
This isn’t just OpenAI’s problem. Earlier this month, the National Cyber Security Centre in the U.K. warned that prompt injection attacks against generative AI systems may never be fully mitigated.
FAKE AI CHAT RESULTS ARE SPREADING DANGEROUS MAC MALWARE
Prompt injection attacks exploit trust at scale, allowing malicious instructions to influence what an AI agent does without the user ever seeing it. (Kurt “CyberGuy” Knutsson)
The risk trade-off with AI browsers
OpenAI says it views prompt injection as a long-term security challenge that requires constant pressure, not a one-time fix. Its approach relies on faster patch cycles, continuous testing, and layered defenses. That puts it broadly in line with rivals like Anthropic and Google, which have both argued that agentic systems need architectural controls and ongoing stress testing.
Where OpenAI is taking a different approach is with something it calls an “LLM-based automated attacker.” In simple terms, OpenAI trained an AI to act like a hacker. Using reinforcement learning, this attacker bot looks for ways to sneak malicious instructions into an AI agent’s workflow.
The bot runs attacks in simulation first. It predicts how the target AI would reason, what steps it would take and where it might fail. Based on that feedback, it refines the attack and tries again. Because this system has insight into the AI’s internal decision-making, OpenAI believes it can surface weaknesses faster than real-world attackers.
Even with these defenses, AI browsers aren’t safe. They combine two things attackers love: autonomy and access. Unlike regular browsers, they don’t just display information, but also read emails, scan documents, click links and take actions on your behalf. That means a single malicious prompt hidden in a webpage, document or message can influence what the AI does without you ever seeing it. Even when safeguards are in place, these agents operate by trusting content at scale, and that trust can be manipulated.
THIRD-PARTY BREACH EXPOSES CHATGPT ACCOUNT DETAILS
As AI browsers gain more autonomy and access to personal data, limiting permissions and keeping human confirmation in the loop becomes critical for safety. (Kurt “CyberGuy” Knutsson)
7 steps you can take to reduce risk with AI browsers
You may not be able to eliminate prompt injection attacks, but you can significantly limit their impact by changing how you use AI tools.
1) Limit what the AI browser can access
Only give an AI browser access to what it absolutely needs. Avoid connecting your primary email account, cloud storage or payment methods unless there’s a clear reason. The more data an AI can see, the more valuable it becomes to attackers. Limiting access reduces the blast radius if something goes wrong.
2) Require confirmation for every sensitive action
Never allow an AI browser to send emails, make purchases or modify account settings without asking you first. Confirmation breaks long attack chains and gives you a moment to spot suspicious behavior. Many prompt injection attacks rely on the AI acting quietly in the background without user review.
3) Use a password manager for all accounts
A password manager ensures every account has a unique, strong password. If an AI browser or malicious page leaks one credential, attackers can’t reuse it elsewhere. Many password managers also refuse to autofill on unfamiliar or suspicious sites, which can alert you that something isn’t right before you manually enter anything.
Next, see if your email has been exposed in past breaches. Our #1 password manager (see Cyberguy.com) pick includes a built-in breach scanner that checks whether your email address or passwords have appeared in known leaks. If you discover a match, immediately change any reused passwords and secure those accounts with new, unique credentials.
Check out the best expert-reviewed password managers of 2025 at Cyberguy.com
4) Run strong antivirus software on your device
Even if an attack starts inside the browser, antivirus software can still detect suspicious scripts, unauthorized system changes or malicious network activity. Strong antivirus software focuses on behavior, not just files, which is critical when dealing with AI-driven or script-based attacks.
The best way to safeguard yourself from malicious links that install malware, potentially accessing your private information, is to have strong antivirus software installed on all your devices. This protection can also alert you to phishing emails and ransomware scams, keeping your personal information and digital assets safe.
Get my picks for the best 2025 antivirus protection winners for your Windows, Mac, Android & iOS devices at Cyberguy.com
5) Avoid broad or open-ended instructions
Telling an AI browser to “handle whatever is needed” gives attackers room to manipulate it through hidden prompts. Be specific about what the AI is allowed to do and what it should never do. Narrow instructions make it harder for malicious content to influence the agent.
6) Be careful with AI summaries and automated scans
When an AI browser scans emails, documents or web pages for you, remember that hidden instructions can live inside that content. Treat AI-generated actions as drafts or suggestions, not final decisions. Review anything the AI plans to act on before approving it.
7) Keep your browser, AI tools and operating system updated
Security fixes for AI browsers evolve quickly as new attack techniques emerge. Delaying updates leaves known weaknesses open longer than necessary. Turning on automatic updates ensures you get protection as soon as they’re available, even if you miss the announcement.
CLICK HERE TO DOWNLOAD THE FOX NEWS APP
Kurt’s key takeaway
There’s been a meteoric rise in AI browsers. We’re now seeing them from major tech companies, including OpenAI’s Atlas, The Browser Company’s Dia, and Perplexity’s Comet. Even existing browsers like Chrome and Edge are pushing hard to add AI and agentic features into their current infrastructure. While these browsers can be useful, the technology is still early. It’s best not to fall for the hype and to wait for it to mature.
Do you think AI browsers are worth the risk today, or are they moving faster than security can keep up? Let us know by writing to us at Cyberguy.com
Sign up for my FREE CyberGuy Report
Get my best tech tips, urgent security alerts, and exclusive deals delivered straight to your inbox. Plus, you’ll get instant access to my Ultimate Scam Survival Guide — free when you join my CYBERGUY.COM newsletter.
Copyright 2025 CyberGuy.com. All rights reserved.
Technology
Google’s Nest Thermostat has hit its best price of the year
If you’re looking for a relatively affordable way to cut down on cooling costs, Google’s Nest Thermostat can help. It’s packed with smart controls and energy-saving features, and right now it’s on sale in white for $79 ($50 off), which is its best price of the year, at Amazon.
The smart thermostat is quick to install and makes it easy to adjust your home’s temperature whether you’re relaxing in bed or on your way home thanks to the Google Home app. You can also create schedules and control it with your voice using Google Assistant, Alexa, or another Matter-compatible voice assistant.
Once it’s set up, the Nest Thermostat can automatically turn the temperature down when you’re away to help reduce unnecessary energy use, while Google’s Savings Finder feature suggests additional ways to save over time. It also monitors your HVAC system and can alert you if something doesn’t seem right, making it easier to stay on top of maintenance before small issues become bigger, more expensive ones. If you’re eligible, Nest Renew can also automatically shift some of your heating and cooling to times when electricity is cleaner or cheaper.
That said, this is Google’s entry-level model from 2020, so you do miss out on some of the premium features found on the latest Nest Learning Thermostat. Unlike the flagship version, it won’t learn your schedule automatically over time, for example, and lacks support for Nest Temperature Sensors that let you prioritize the temperature in a specific room. Even so, if all you want is an easy way to adjust your home’s temperature remotely and potentially lower your energy bills, the Nest Thermostat is still a solid investment at this price.
Technology
Medical identity theft follows you into the doctor’s office
NEWYou can now listen to Fox News articles!
The Justice Department recently charged 455 people in its annual National Health Care Fraud Takedown. The cases involve more than $6.5 billion in alleged false claims. More state Medicaid units took part than in any prior year. Ninety of the accused are doctors or other licensed medical professionals. The DOJ says prosecutors still must prove the charges in court.
Many schemes used other people’s medical identities. Prosecutors also added aggravated identity theft charges in cases across dozens of states. In one case, the co-owner of a Virginia mental health company allegedly paid homeless people with hotel stays. Prosecutors say the company used their Medicaid numbers, then billed Medicaid for crisis services the patients never got.
For the people whose numbers got used, the case file may eventually close. Their medical records may not be so easy to fix. Once someone else’s treatment shows up under your name, it can add wrong information to your chart. It can also use up insurance benefits you may need later. That is harder to undo than canceling a credit card.
Sign up for my FREE CyberGuy Report
- Get my best tech tips, urgent security alerts and exclusive deals delivered straight to your inbox.
- For simple, real-world ways to spot scams early and stay protected, visit CyberGuy.com – trusted by millions who watch CyberGuy on TV daily.
- Plus, you’ll get instant access to my Ultimate Scam Survival Guide free when you join.
DR OZ WARNS MEDICARE SCAMMERS ARE STEALING BILLIONS — AND YOUR PERSONAL INFORMATION COULD BE NEXT
Medical identity theft can put someone else’s claims, prescriptions or diagnoses into your health records, creating problems that can follow you into a doctor’s office. (iStock)
The identity thief’s treatment gets written into your file
Medical identity theft happens when someone uses your name, Social Security number (SSN), health insurance account number, or Medicare number to see a doctor, fill a prescription, buy medical equipment, or submit a claim, according to the Federal Trade Commission (FTC).
When care is billed under your name, the thief’s health information can blend into yours. The FTC warns that mixed records can affect the care you’re able to get and the benefits you are able to use. A blood type, a drug allergy, a diagnosis, or a prescription that belongs to a stranger can sit in the file a physician reads before treating you.
Data breaches can feed the market for medical identity theft
Hospitals and insurers hold the exact records that make the fraud work, and those records are stolen often. This does not mean every healthcare breach leads to fraud. However, it explains why your insurance number, Medicare number, SSN and medical records can become valuable long after a breach notice arrives.
This spring, NYC Health + Hospitals reported that an intruder had copied files that may have included health insurance information, medical information, biometric data, billing data and other personal information. The breach was later reported to affect roughly 1.8 million current and former patients and employees.
Once a name, SSN, insurance number, Medicare number or medical record reaches a criminal marketplace, it can be resold to operators who bill under someone else’s identity.
Treat your insurance card like a credit card
Your health insurance and Medicare numbers are what these operations need, so the FTC recommends guarding them the way you would a payment card.
- Keep enrollment forms, benefit statements, and prescription labels somewhere secure, and shred them before throwing them out.
- When a doctor’s office asks for your SSN, ask whether it can use another identifier or the last four digits instead.
- Be wary of anyone who calls, texts, or emails offering free braces, genetic tests, or medical supplies in exchange for your Medicare number; several of the schemes in the June takedown billed Medicare for exactly those items.
- If you are on Medicare, create or log in to your secure Medicare account and review your claims. You can also check your Medicare Summary Notice for services, supplies or equipment you do not recognize. If something looks wrong, call 1-800-MEDICARE.
HOSPICE FRAUD USES STOLEN IDENTITIES FOR FAKE PATIENTS
Experts urge patients to treat insurance cards like credit cards and quickly challenge unfamiliar medical bills, claims or benefits notices. (iStock)
Your credit report may never flag this fraud
Because a fraudulent medical claim runs through insurance and provider systems instead of a credit check, it skips the alerts most people rely on.
Here’s what the FTC says you should look out for:
- A bill or an Explanation of Benefits (EOB) statement for care you never received
- A call from a debt collector about a medical debt you do not owe
- A medical collection you do not recognize on your credit report
- A notice from your insurer that you have reached your benefit limit
- A Medicare Summary Notice that lists services, supplies or equipment you never received
What to do first if a medical claim looks wrong
If a bill, EOB or Medicare notice shows care you never received, move quickly and keep everything in writing.
1) Call your insurer or Medicare directly
Call your insurer or Medicare using the number on your card, not a number from a random text, email or voicemail.
2) Get the claim details
Ask for the provider name, date of service, claim number and service details.
3) Request the records in writing
Contact the provider in writing and request the medical or billing records tied to that claim.
4) Report the error
Report the error to your insurer’s fraud department.
5) File an identity theft report
File a report at IdentityTheft.gov if your medical identity was used. That gives you a recovery plan and documentation you may need if fraudulent bills or collections show up later.
6) Save every document
Keep copies of every bill, EOB, letter, portal message, police report and case number.
Correcting a medical file is slower than disputing a charge
Request your records from every provider, clinic, pharmacy, lab and insurer the thief may have used, then report each error in writing. Under HIPAA, a provider generally has 30 days to give you access to your records after a written request, with a possible 30-day extension.
Fixing the record itself can take longer. HHS says a covered provider or health plan usually has up to 60 days to act on a request to amend a medical record, with a possible 30-day extension in certain cases. If the provider or plan created the wrong information, it must amend inaccurate or incomplete information.
There’s one catch, though: a provider may refuse to release records that now contain a stranger’s information, citing that person’s privacy. If that happens, ask for the provider’s privacy officer or patient advocate. You can also file a complaint with the U.S. Department of Health and Human Services Office for Civil Rights if you do not get your records or an explanation within the required window.
TEXAS DATA BREACH HITS 3M LICENSE CUSTOMERS
Stolen Medicare, Medicaid or insurance numbers can be used to bill for care, medical equipment or prescriptions patients never received. (kali9/Getty Images)
A credit freeze alone won’t stop a claim under your insurance
A freeze blocks new accounts, but it does nothing about a claim filed with your insurance number. Because medical identity theft can move without touching your credit file, monitoring where your personal information appears is the earliest way to act on it.
An identity theft protection service can monitor the dark web, data broker sites and people-search sites for exposed SSNs, driver’s license numbers, medical ID numbers and email addresses. It can also track all three credit bureaus for medical collections that may follow and flag public-record changes tied to your name.
If misuse happens, some services include fraud resolution support to help you request records, dispute fraudulent claims and work with providers, insurers and credit bureaus. Some plans also include identity theft insurance for eligible recovery costs.
No service can prevent every misuse of your medical identity. However, ongoing monitoring may flag exposed information before another person’s treatment reaches your records and your insurance.
See my tips and best picks on Best Identity Theft Protection at CyberGuy.com.
Kurt’s key takeaways
Medical identity theft hits in a place most of us rarely check: our health records. A stolen credit card can usually be canceled quickly. A stolen Medicare or insurance number can create fake claims, wrong diagnoses and benefit headaches that follow you long after the fraud case ends. I would not wait for a credit alert here. Check your EOBs, Medicare Summary Notices and insurer portals for visits, prescriptions or equipment you never received. Also, treat your insurance card like a payment card. Do not give the number to anyone who calls, texts or emails out of nowhere with a free offer. The most important thing is to act fast. Call your insurer or Medicare, ask for the claim details and request your medical records in writing. Then file at IdentityTheft.gov, so you have the paperwork you need if fraudulent bills or collections show up later.
Have you ever spotted a medical bill, insurance claim or EOB for care you never received? Let us know by writing to us at CyberGuy.com.
CLICK HERE TO DOWNLOAD THE FOX NEWS APP
Sign up for my FREE CyberGuy Report
- Get my best tech tips, urgent security alerts and exclusive deals delivered straight to your inbox.
- For simple, real-world ways to spot scams early and stay protected, visit CyberGuy.com – trusted by millions who watch CyberGuy on TV daily.
- Plus, you’ll get instant access to my Ultimate Scam Survival Guide free when you join.
Copyright 2026 CyberGuy.com. All rights reserved.
Technology
Meta is reportedly working on smart glasses that would be recording all the time
Meta might be the next company to make an always-on AI wearable. The company is working on prototype “super sensing” always-aware smart glasses that could continuously record audio and snap photos “every few seconds,” according to the Financial Times. The wearer could then ask Meta AI about the captured audio and images.
However, the images and audio might not be directly available to the user. Here’s how the FT describes one way the glasses could use the data:
In one proposed system, raw footage and audio would not be stored by Meta or made available to the user, several people said. Instead, the metadata from that audio and images would be extracted and uploaded to the server for Meta’s AI to query, which proponents argue would have fewer privacy implications.
But currently, Meta is planning for the LED recording indicator to remain off in “super sensing” mode, the FT reports. In a July 2025 whitepaper, the company said that it would reserve the LED indicator for “active capture” scenarios where the user is saving photos or videos, and leave it off during “AI Feature” use — such as scanning a menu — to avoid users becoming too used to the indicator. (If the indicator was on during the “super sensing” mode, it might also be harder to know when the glasses are actually recording video.)
Meta is also discussing if it would use the captured data for training its AI models. It may also bring the “super sensing” features to glasses it has already released, the FT says.
“While we don’t comment on internal prototypes, we’re committed to getting our glasses right because they need to be loved by both people wearing them and those around them,” Meta spokesperson Dave Arnold says in a statement to The Verge. Arnold also notes that “Our approach has been to develop new technologies that will help people throughout their day, with privacy built in from the ground up.”
Meta hasn’t been shy about some type of always-aware glasses being a possibility. CEO Mark Zuckerberg, in the company’s Q1 2026 earnings call, said that he was “really excited to see the glasses evolve from being able to answer questions to being able to be a personal agent that’s with you all day long, helping you remember things and achieve your goals.” In a March blog post about new Ray-Ban Meta glasses, the company wrote that “with ongoing software updates, Meta AI on glasses will transition from something you have to prompt with a question each time, to a more continuous, in-the-moment assistant that can help throughout the day.”
-
Business5 minutes agoWaymo is starting robotaxi service in San Diego
-
Entertainment8 minutes ago‘Children of Blood and Bone’ author won’t see film after feud with star Amandla Stenberg
-
Lifestyle13 minutes agoAfter her son’s death, she found a new purpose. ‘He’s whispering: Mom, this is your path’
-
Politics20 minutes agoIran ceasefire is ‘over,’ Trump says, and orders additional strikes
-
Science23 minutes agoDiarrhea-causing cyclosporiasis exceeds 1,000 cases in U.S. What Californians should know
-
Sports28 minutes agoArthur Fery’s fairy-tale Wimbledon run puts British wild card on brink of history
-
World38 minutes agoBurnham on course to become next UK PM with backing of 322 Labour MPs
-
News1 hour agoTrump fires last members of election commission, inciting fears of midterm ‘chaos’