Nerd Free AI beaver mascot

Nerd Free AI

AI news and how-tos, minus the jargon

Author: Nerd Free AI Team

  • Big Websites Are Blocking AI Agents. Here’s What That Means If You Use One.

    Big Websites Are Blocking AI Agents. Here’s What That Means If You Use One.

    The short version: On Friday, October 17, 2026, The Verge reported that a growing list of popular websites are blocking AI agents, the new AI helpers that can go onto websites and do tasks for you, like booking a flight or ordering groceries. Delta Air Lines won’t let them book flights. JPMorgan Chase and Charles Schwab are keeping them out of customer accounts. Yelp only lets in agents whose makers pay up. Here’s what’s going on, in plain English, and what it means if you use one (or are thinking about it).

    First, what’s an AI agent?

    A regular chatbot answers questions. An agent goes a step further: it actually does things on websites for you. You might say “get me a table for four on Friday at 7” or “reorder paper towels,” and the agent opens websites, clicks buttons, and fills in forms on its own. Think of a very fast personal assistant who never needs a coffee break.

    The biggest tech companies are betting heavily on them. Meta has Muse, ChatGPT has an agent called Dots, Elon Musk’s xAI has Grok, and a startup called Instinct has its own. (We explained Meta’s newest Muse AI here.)

    Who’s closing the door

    • Delta Air Lines. Delta told TechCrunch it has no partnership that lets an outside AI agent shop for or book Delta flights for customers, and that it’s taking steps to protect customers from “unauthorized automated activity.” It says it’s still looking at how agents could help in the future.
    • United Airlines. United pointed TechCrunch to its website rules, which say you can’t use any “robot, spider, other automatic device” on its site without written permission.
    • JPMorgan Chase and Charles Schwab. Both confirmed to The Verge that they’re blocking AI agents from getting into customer accounts because of security concerns.
    • Yelp. Yelp only allows AI agents whose makers pay for its data. So an agent trying to book you a table or put you on a waitlist through Yelp will probably fail.
    • Amazon. Amazon has fought AI shopping agents for a while. Last year it banned the shopping agent from an AI company called Perplexity and sued it. Now Amazon says Meta’s Muse breaks its site rules, and users report Grok, Instinct and Dots can’t use Amazon either.
    • Others, according to users. People say Cars.com and Amtrak recently stopped working with their agents, and complaints online have named eBay, Zillow, Pizza Hut and Adidas. Those are user reports, not confirmed bans. eBay told TechCrunch it doesn’t ban all shopping agents, just unauthorized ones, and Zillow said it allows agents working for customers but its anti-bot defenses sometimes catch them by mistake.

    Why websites are doing this

    There isn’t just one reason, and not all of them are bad for you.

    • Security. Your bank has no easy way to tell your helpful agent apart from a crook’s bot trying to break into accounts. Keeping all bots out is the safer bet.
    • Money. Perplexity argued that Amazon doesn’t want agents shopping because fewer human eyes on the site means fewer people seeing Amazon’s ads. Yelp, meanwhile, wants AI companies to pay for its information.
    • Accidents. Lots of websites use “prove you’re human” checks to stop spammers and scammers, and AI agents set them off too. Walmart actually partners with Meta’s Muse, but some shoppers still got booted at a “verify you’re human” button. Walmart told TechCrunch that wasn’t on purpose.

    What’s being done about it

    Meta’s take: “Turning away a personal agent means turning away the customer behind it.” Instead of fighting, Meta has teamed up with stores like Walmart, Gap, Sephora and Best Buy so Muse is welcome there. Shopify, which runs the online stores for lots of businesses, has opened its shops to agents too.

    Meta, Walmart, the payment company Stripe and several other companies are also working on an open standard, which is a shared set of rules, so websites can tell good agents (working for a real customer) apart from bad bots. It isn’t finished yet.

    One expert sees a downside for everyone. Chris Canal, who runs the AI safety testing company EquiStamp, told The Verge that all these blocks could make the web more “annoying,” with more “prove you’re human” hoops to jump through, maybe even on a simple grocery website.

    What this means for you

    • If your agent gets stuck, it’s probably not you. The website may be blocking it on purpose or by accident. Finish the job yourself on the company’s official website or app.
    • Stick to partner stores. Some agents show a list of stores they officially work with (Muse calls them “connectors”). Those are the least likely to break halfway through an order.
    • Keep agents away from your bank. Your bank may block them anyway, and giving any app your banking password is risky. If your bank turns an agent away, that’s the bank doing its job.
    • Protect accounts you care about. A few people told TechCrunch that eBay suspended their accounts after they used an AI agent there. Check a site’s rules before letting an agent loose on an account you’d hate to lose.
    • Always check before you pay. Agents can get confused. Look over the dates, sizes, prices and final total yourself before anything gets bought.
    • Expect more “I’m not a robot” checks. They’re a pain, but they also help keep scammers out. And if a text or email says your AI agent needs your password to “fix” an order, that’s a red flag. Here’s how to ask a free AI chatbot if a text is a scam.

    New to chatbots? Our Try It Yourself guides walk you through easy, practical ways to use free AI, step by step.


    Common questions

    What is an AI agent?

    An AI agent is an AI helper that does tasks for you on websites, like booking a flight, making a reservation, or buying something, instead of just answering questions. Examples include Meta’s Muse, ChatGPT’s Dots, xAI’s Grok, and Instinct.

    Why won’t Delta let my AI agent book a flight?

    Delta says it doesn’t have a partnership that lets outside AI agents shop for or book Delta flights, and that it’s protecting customers from unauthorized automated activity. For now, book Delta flights yourself on Delta’s website or app.

    Is it safe to let an AI agent into my bank account?

    Big banks like JPMorgan Chase and Charles Schwab are blocking AI agents from customer accounts because of security concerns. The safest choice is to do your own banking on your bank’s official app or website, and never share your banking password with another app.

    Will AI agents ever work on every website?

    Nobody knows yet. Companies including Meta, Walmart and Stripe are working on shared rules to help websites trust good agents, and some stores have signed partnerships. For now, agents work best at stores that officially partner with them.

    Want tech that helps without the headaches?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks


    Sources

  • Make a Last-Minute Halloween Costume from Stuff You Already Own, with Free AI

    Make a Last-Minute Halloween Costume from Stuff You Already Own, with Free AI

    Halloween is Saturday, October 31, and costumes from the store add up fast, especially if you’re dressing a whole family. Meanwhile, your closet is full of clothes that could become something fun, if only you had an idea.

    That’s where a free AI chatbot comes in. Tell it what you’ve got, or even show it a photo, and it will suggest costumes you can make from things you already own, walk you through putting one together, and make a tiny shopping list for anything missing. That old flannel shirt might be one cardboard axe away from lumberjack greatness.

    What you need

    • A phone, tablet, or computer with internet.
    • A free AI chatbot. Any of these work: ChatGPT (chatgpt.com), Google Gemini (gemini.google.com), or Microsoft Copilot (copilot.microsoft.com). Each also has a free phone app.
    • A few minutes to dig through your closet, junk drawer, craft supplies, or recycling bin.
    • Optional: your phone’s camera, so the AI can see what you have.
    • About 20 to 30 minutes for ideas and a plan (making the costume takes longer, of course).

    Step 1: Tell the AI who the costume is for

    Start with the basics: who’s wearing it, what they like, and how much time and money you want to spend. The more you share, the better the ideas. Copy this prompt and fill in the brackets:

    I need a homemade Halloween costume, and I want to use things I already own.
    It's for: [me / my 8-year-old / my partner and me / my dog]
    They love: [for example, dinosaurs, old movies, cooking, sports]
    Where it will be worn: [trick-or-treating / a work party / handing out candy at home]
    Time to make it: [one evening / 30 minutes]
    Budget for anything extra: [for example, $10]
    Give me 8 fun ideas. For each one, list what I'd probably need and how hard it is (easy, medium, or hard).

    Step 2: Show it what you’ve got

    This is the fun part. Lay out a few possible costume pieces on your bed or the floor: shirts, hats, scarves, old Halloween leftovers, cardboard boxes, anything. Then snap a photo. Most free chatbots let you add a picture: look for a +, a paperclip, or a camera button next to the spot where you type. (Free plans may limit how many photos you can add each day.)

    Here's a photo of clothes and stuff I have at home.
    What Halloween costumes could I make mostly from these things?
    Give me 5 ideas, from easiest to most creative.
    For each one, tell me which items in the photo to use and what else I'd need, if anything.

    No camera button, or just don’t feel like taking a picture? No problem. Type a quick list instead, like “black turtleneck, gray sweatpants, a yellow raincoat, cardboard boxes, face paint from last year.” It works just as well.

    Step 3: Pick a favorite and get the step-by-step

    Found an idea you like? Ask for simple instructions. If you’re not the crafty type, say so. The AI will happily skip the sewing machine.

    Let's go with the [costume idea] costume.
    Give me simple step-by-step instructions to make it.
    I don't sew, so use things like safety pins, tape, glue, cardboard, and markers.
    Tell me which steps an adult should do (like cutting or hot glue) and which ones a kid can help with.
    It needs to be comfortable to walk around in for a couple of hours.

    If a step doesn’t make sense, just ask: “Can you explain step 3 more simply?” You can keep asking follow-up questions as long as you like.

    Step 4: Go bigger with a group or family theme

    Dressing up as a group is twice the fun and often easier, because everyone can share materials. Try this:

    Suggest 5 group Halloween costume themes for [number] people: [for example, two adults, a 6-year-old, and a toddler in a stroller].
    Use things most families already have at home.
    For each theme, give every person a role and a simple costume idea.
    Keep it family-friendly and not too scary for little kids.

    Step 5: Make a tiny shopping list for anything missing

    Sometimes you’re just one piece short, like a pair of cat ears or a cheap mustache. Ask the AI to keep the list short and cheap:

    For the [costume idea] costume, what am I still missing?
    Make a short shopping list of only the things I don't have.
    Suggest cheap places to look, like a dollar store, thrift store, or craft store, and anything I could borrow from a friend or make from cardboard instead.
    My budget is [amount].

    The AI doesn’t know what’s on the shelf at your local store or today’s prices, so treat its list as a starting point, not a guarantee.

    Step 6: Add the finishing touches

    Makeup, props, and a way to be seen in the dark take a costume from “nice try” to “whoa.” Ask for simple ideas that are safe and comfortable:

    Give me ideas to finish the [costume idea] costume:
    1. A simple face paint look a beginner can do in 10 minutes (no masks that cover the eyes).
    2. A soft, lightweight prop that's safe to carry around.
    3. Ways to add reflective tape or lights so we're easy to see after dark, without ruining the look.

    Before the big night, try the whole costume on and walk around the house for a few minutes. It’s much better to find the wobbly hat or the pinching shoe on Tuesday than on Halloween night.

    Safety tips

    • Be easy to see. Add reflective tape to costumes and treat bags, and carry a flashlight or glow stick after dark.
    • Watch out for flames. For store-bought pieces like wigs and capes, look for a “flame resistant” label, and keep long, flowing fabric away from candles and jack-o’-lanterns.
    • Pick face paint over masks. Masks can block your view. If you use face paint or makeup, test a small spot on your arm a day or two ahead, and only use products made for skin, not craft paint.
    • Skip colored contact lenses without a prescription. The FDA warns that decorative lenses bought without an eye exam can cause infections and even vision loss.
    • Keep it short and soft. Hem or pin up anything that could trip you, and choose short, bendy props instead of sharp swords or long sticks.
    • AI ideas aren’t safety-checked. If a chatbot suggests something that seems risky, like a real candle in a costume or hot glue near skin, skip it and ask for a safer option.
    • Mind what’s in your photos. Pictures of clothes are fine. Leave out faces, mail with your address, or anything else personal, and check your chatbot’s privacy settings if you’d rather it not save your chats.

    Try this next

    Throwing a little Halloween get-together? Our guide to cheap dinners from what’s in your fridge works great for party snacks too. And if the holidays are starting to look expensive, here’s how to make a simple monthly budget with free AI.

    Common questions

    Can a free AI chatbot look at a photo of my clothes?

    Usually, yes. Free versions of ChatGPT, Google Gemini, and Microsoft Copilot let you add a photo by tapping a plus sign, paperclip, or camera button, though free plans may limit how many you can add each day. If yours can’t, type a list of what you have instead.

    What if I’m not crafty at all?

    Tell the AI that. Ask for “no-sew, no-glue costume ideas I can put together in 15 minutes from regular clothes,” and it will stick to easy ideas like a quick character outfit or a simple pun costume.

    Is face paint safe to use?

    Use makeup made for skin, not craft paint, and test a small spot on your arm a day or two before Halloween. Wash it off gently when you’re done, and don’t share makeup.

    When is Halloween this year?

    Halloween 2026 is on Saturday, October 31.

    Want gadgets that make everyday life easier?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks


    Sources

  • Compare Health Insurance Plans During Open Enrollment with Free AI

    Compare Health Insurance Plans During Open Enrollment with Free AI

    It’s open enrollment season, the once-a-year window to pick or change your health insurance. And every plan comes with a pile of confusing words: deductibles, coinsurance, out-of-pocket maximums. It’s enough to make anyone pick the same plan as last year and hope for the best. (No judgment. We’ve all done it.)

    Good news: a free AI chatbot can help. It can explain the jargon in plain English, line up your choices side by side, and estimate what each plan could cost you in a real year. You still make the choice. The AI just helps you see it clearly.

    Key dates this year

    • Medicare: October 15 to December 7, 2026. Changes start January 1, 2027.
    • HealthCare.gov (the Marketplace): November 1, 2026 to January 15, 2027. Sign up by December 15 for coverage that starts January 1. A few states run their own sites with different deadlines.
    • Insurance through your job: Dates vary by employer, and many hold open enrollment in the fall. Check your benefits website or ask HR.

    What you need

    • A phone, tablet, or computer with internet.
    • A free AI chatbot. Any of these work: ChatGPT (chatgpt.com), Google Gemini (gemini.google.com), or Microsoft Copilot (copilot.microsoft.com). Each also has a free phone app.
    • The details of the plans you’re choosing between (Step 1 shows where to find them).
    • A list of your doctors and the prescriptions you take.
    • A rough idea of how often you went to the doctor this year.
    • About 30 to 45 minutes.

    Step 1: Gather the plan details

    First, find the plans you can choose from. Where you look depends on how you get insurance:

    • Through your job: Check your benefits website or ask HR. Look for each plan’s Summary of Benefits and Coverage, a short, standard document that health plans have to give you. It lists the key costs in the same format for every plan, which makes comparing easier.
    • On the Marketplace: You can preview plans and prices on HealthCare.gov before open enrollment starts, without signing up.
    • On Medicare: Use Medicare’s Plan Finder. If you already have a plan, also look for its Annual Notice of Change, a letter your plan sends each fall listing what’s changing next year.

    For each plan, write down (or copy) the key numbers: the monthly cost, the deductible, what a doctor visit costs, what your prescriptions cost, and the out-of-pocket maximum. Don’t worry if those words mean nothing yet. That’s Step 2.

    Step 2: Get the jargon in plain English

    Insurance has its own language. Here’s a quick cheat sheet:

    • Premium: what you pay every month just to have the insurance, even if you never see a doctor.
    • Deductible: what you pay for care yourself each year before the plan starts paying its share. (Many plans cover some things, like yearly checkups, even before you hit it.)
    • Copay: a flat fee for a visit or prescription, like $30.
    • Coinsurance: your share of the bill as a percentage, like 20%, usually after you’ve met your deductible.
    • Out-of-pocket maximum: the most you’d pay for covered, in-network care in a year. After that, the plan pays 100%. Your monthly premiums don’t count toward it.
    • Network: the doctors, hospitals, and pharmacies that have a deal with the plan. Going “out of network” usually costs a lot more.

    Seeing a word that isn’t on the list? Medicare uses a few extra terms, like “Part D” (drug coverage) and “Medicare Advantage.” Copy this prompt and fill in any words that confuse you:

    I'm choosing a health insurance plan and some of the words confuse me.
    Please explain these in plain English, like you're talking to a friend: [list the words, like deductible, coinsurance, formulary].
    For each one, give a simple example with dollar amounts.
    Keep it short.

    Step 3: Line up your plans side by side

    Now let the AI build a comparison table. Type in each plan’s numbers, or copy and paste the text from its summary. Leave out anything with your name, member ID, or other personal details.

    Please put these health plans side by side in a simple table.
    Rows: monthly premium, total premiums for the year, deductible, doctor visit cost, specialist visit cost, generic drug cost, emergency room cost, and out-of-pocket maximum.
    Then explain the biggest differences in plain English.
    Only use the numbers I give you. If something is missing, write "not listed" instead of guessing.
    
    Plan A: [paste or type the details]
    Plan B: [paste or type the details]
    Plan C (optional): [paste or type the details]

    That last instruction matters. When a chatbot doesn’t have a number, it will sometimes fill the gap with a confident-sounding guess. Telling it not to guess helps keep the table honest.

    Step 4: Play “what if” with a quiet year and a rough year

    Here’s the big secret of open enrollment: the plan with the cheapest monthly price isn’t always the cheapest plan. It depends on how much care you end up needing. Look at these two made-up plans:

    Plan APlan B
    Monthly premium$150$60
    Deductible$500$3,000
    Out-of-pocket maximum$3,000$7,000
    Quiet year (premiums only, plus a little for a checkup or two)About $1,800About $720
    Rough year (premiums plus hitting the out-of-pocket maximum)About $4,800About $7,720

    In a quiet year, Plan B saves you about $1,000. In a rough year, like a surgery or a hospital stay, Plan A saves you almost $3,000. Neither one is “right.” It depends on your health, your budget, and how much risk you’re comfortable with.

    Ask the AI to run the numbers for your plans:

    Using the plans in the table, estimate my total cost for one year (premiums plus what I'd pay for care) in three situations:
    1. A quiet year: [for example, one checkup and two sick visits]
    2. A typical year for me: [describe it, like 6 doctor visits, 2 specialist visits, and one generic prescription every month]
    3. A rough year: a hospital stay or surgery where I reach the out-of-pocket maximum.
    Show your math step by step for each plan.
    Then tell me which plan costs the least in each situation.

    Chatbots are great at explaining but can still slip up on simple math. Check the big totals yourself with your phone’s calculator. It takes a minute and could save you hundreds.

    Step 5: Check your doctors and prescriptions (don’t trust the AI on this one)

    A cheap plan isn’t a deal if your favorite doctor isn’t in it. AI chatbots don’t reliably know which doctors are in a plan’s network or which drugs a plan will cover next year, and those lists change. So check with the real sources:

    • Doctors: Use the plan’s online “find a doctor” tool, then call your doctor’s office to confirm. Office staff hear this question all the time.
    • Prescriptions: Look up each drug on the plan’s drug list (sometimes called a “formulary”). On Medicare’s Plan Finder, add your drugs and pharmacy to see estimated yearly costs for each plan.

    Nervous about making the call? Have the AI write you a script:

    Write me a short, polite script for calling my doctor's office.
    I want to ask if they will accept [insurance company and exact plan name] next year.
    Include 3 questions to ask, and a spot to write down the name of the person I talked to and the date.

    Step 6: Make a question list, then get free help from a real person

    Before you decide, have the AI turn everything into a short to-do list of questions:

    Based on everything we talked about, make a short checklist of questions I should ask [HR / the insurance company / a Medicare counselor] before I choose a plan.
    Put the most important questions first.

    Then take your list to a real person. This help is free and isn’t trying to sell you anything:

    • Medicare: Your local State Health Insurance Assistance Program (SHIP) gives free, one-on-one Medicare counseling. Find yours at shiphelp.org, or call 1-800-MEDICARE (1-800-633-4227).
    • Marketplace: Find free, trained helpers near you at localhelp.healthcare.gov.
    • Job-based insurance: Your HR or benefits team can answer questions about your company’s plans.

    Finally, pick your plan and enroll before the deadline. If you do nothing, many plans simply renew for next year, but the prices and coverage can change. That’s exactly why it pays to take a look.

    Safety tips

    • Never share personal ID numbers. No Social Security number, Medicare number, member ID, or date of birth. Plan names and dollar amounts are all the AI needs.
    • Keep health details general. “One generic prescription a month” works just as well as naming your condition.
    • Double-check every number. Compare the AI’s table against the official plan documents. AI can misread or mix up numbers.
    • AI isn’t an insurance expert. The final word on what’s covered comes from the plan itself, HealthCare.gov, Medicare.gov, or your HR team.
    • Watch out for open enrollment scams. Real Medicare won’t call you out of the blue to sell you a plan or ask for your Medicare number. Got a suspicious text? See how to ask AI if a text is a scam.
    • Check your chatbot’s privacy settings. Many let you turn off chat history or stop your chats from being used to train their AI. In ChatGPT, a Temporary Chat isn’t saved to your history. Curious how ChatGPT uses your chats for ads? Read our plain-English explainer.

    Try this next

    Once you’ve picked a plan, our guide to getting ready for a doctor’s visit with free AI helps you make the most of your appointments. And if a confusing bill or insurance letter shows up later, here’s how to get AI to explain it in plain English.

    Common questions

    Can AI pick my health insurance plan for me?

    No, and it shouldn’t. A chatbot is great for explaining terms, building a comparison table, and estimating costs, but it can make mistakes and doesn’t know every plan’s network or drug list. Use it to understand your options, then confirm with the plan or a free counselor.

    Is it safe to put my insurance information into a chatbot?

    Plan names and dollar amounts are fine. Never share your Social Security number, Medicare number, member ID, date of birth, or other personal details.

    When is open enrollment this year?

    Medicare open enrollment runs October 15 to December 7, 2026. HealthCare.gov open enrollment runs November 1, 2026 to January 15, 2027, and you need to sign up by December 15 for coverage starting January 1. Job-based plans set their own dates.

    Where can I get free help from a real person?

    For Medicare, contact your local SHIP at shiphelp.org or call 1-800-MEDICARE. For Marketplace plans, find free local helpers at localhelp.healthcare.gov. For job-based plans, ask your HR or benefits team.

    Want gadgets that make everyday life easier?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks

  • ChatGPT Ads Can Be Aimed at You Using Your Email or Phone Number. Here’s What You Can Control.

    ChatGPT Ads Can Be Aimed at You Using Your Email or Phone Number. Here’s What You Can Control.

    The short version: On Thursday, October 15, 2026, the news site 404 Media reported that companies advertising inside ChatGPT can aim their ads at groups of people using email addresses, phone numbers, or a special ad ID tied to Android phones. OpenAI’s own help pages for advertisers confirm how it works. Ads only show up for people on ChatGPT’s Free plan and its low-cost Go plan, and there are settings you can change today. Here’s what’s going on, in plain English.

    Wait, ChatGPT has ads now?

    For some people, yes. OpenAI started testing ads in the U.S. on February 9, 2026, and has slowly been turning them on for people using the Free and Go plans. Paid Plus, Pro, Business, Enterprise, and Edu accounts don’t get ads. OpenAI also says it doesn’t show ads to accounts it believes belong to kids under 18.

    The ads appear in a labeled box underneath ChatGPT’s answer. OpenAI is also testing picture ads that pop up while ChatGPT is making an image for you, then disappear when your image is done.

    How ChatGPT decides which ad to show you

    According to OpenAI, the biggest clue is what you’re chatting about right now. Ask for help planning a camping trip, and you might see an ad for tents. It also uses basic info like your general location and language.

    If a setting called personalized ads is turned on, ChatGPT can also use your past chats, the things it remembers about you (its “memory”), and which ads you’ve clicked or hidden before.

    The part most people don’t know: your email and phone number

    This is what 404 Media’s report was about. The news site ran two real ad campaigns in ChatGPT to see how the system works from the advertiser’s side.

    Here’s the idea. A company uploads a list of email addresses, phone numbers, or Android phone ad IDs, usually its customers or people it wants to reach. ChatGPT checks which of those match real ChatGPT accounts. The people who match can then be shown that company’s ads. OpenAI calls these lists custom audiences, and its own instructions for advertisers spell out exactly how to upload one.

    Where would a company get your email or phone number? Lots of places: a store loyalty card, an online order, a newsletter you signed up for, or data companies that buy and sell contact lists. (Remember that 10%-off coupon you traded your email for? Yep, that kind of thing.)

    To be fair, this isn’t brand new. Facebook, Instagram, and Google have let advertisers do the same thing for years, and 404 Media called it “somewhat standard practice” in the ad industry. What’s different is that ChatGPT can also use what you type to pick ads, and plenty of people treat ChatGPT like a private diary.

    The limits OpenAI says it has in place

    • No tiny lists. An advertiser’s list has to match at least 25,000 ChatGPT users before it can be used to target ads. So it isn’t built to single out one person.
    • No peeking at who matched. Advertisers can’t see which people on their list use ChatGPT, and they can’t pick out specific people.
    • Uploaded lists get deleted. OpenAI says the files advertisers upload are deleted after processing, usually within 24 hours.
    • Your chats stay private from advertisers. OpenAI says advertisers never get your chats, chat history, memories, name, email, precise location, or IP address (your internet address). They only see totals, like how many people viewed or clicked an ad.
    • No ads near sensitive topics. OpenAI says ads won’t appear next to chats about personal health, mental health, or politics, and it doesn’t allow political ads.
    • No ads in Temporary Chats. More on that handy feature below.
    • Ads don’t change the answers. OpenAI says ads run on a separate system and advertisers can’t shape what ChatGPT tells you.

    Keep in mind, those are OpenAI’s own promises, and outside experts haven’t checked them. Zach Edwards, an ad-industry researcher quoted by 404 Media, compared OpenAI’s slow rollout of ads to “boiling the frog”: turning up the heat a little at a time so nobody notices.

    What the settings do (and don’t do)

    If you use the Free or Go plan and you’re seeing ads, open ChatGPT’s Settings and look for Ads controls. (Don’t see it on your phone? Update the app.) In there, you can:

    • Turn off “Personalize ads.” ChatGPT stops using your other chats, ad history, and topics to pick ads.
    • Turn off “Past chats and memory.” Your old conversations and ChatGPT’s memory of you won’t be used for ads.
    • See your ad history and topics. This shows which topics ChatGPT thinks you’re interested in.
    • Delete your ads data. OpenAI says it can take up to 30 days to fully disappear from its systems. Your chats aren’t affected.
    • Hide or report a single ad. Tap the three dots on any ad. “About this ad” tells you why you’re seeing it.

    Two catches. First, turning off personalized ads doesn’t turn off ads. You’ll still see ads based on the chat you’re in. Second, it isn’t clear whether that switch keeps you out of those email and phone number lists. Gizmodo asked OpenAI about it and hadn’t heard back when it published its story.

    Don’t want ads at all? In Ads controls, Free users can switch to an Ads-Free option, but you get fewer messages per day and lose some tools, like making images and deep research. Or you can pay for Plus or Pro, which have no ads.

    What this means for you

    • Use Temporary Chat for personal stuff. Temporary Chats show no ads and don’t use or update ChatGPT’s memory. Look for the Temporary Chat option when you start a new chat.
    • Take two minutes to check Ads controls. If you’d rather not have your old chats shaping ads, turn off “Personalize ads” and “Past chats and memory.”
    • An ad is still an ad. OpenAI says ads don’t change ChatGPT’s answers, but the ad itself is a company paying for your attention. Check reviews and compare prices before you buy.
    • Android users can delete their ad ID. On many Android phones, go to Settings, then Privacy, then Ads, and tap “Delete advertising ID.” Menu names vary a little by phone.
    • Watch for fake “deals.” If a text or email claiming to be from ChatGPT offers a too-good-to-be-true deal, be suspicious. Here’s how to ask a free AI chatbot if a text is a scam.

    New to chatbots? Our Try It Yourself guides walk you through easy, practical ways to use free AI, step by step.


    Common questions

    Do paid ChatGPT users see ads?

    No. OpenAI says Plus, Pro, Business, Enterprise, and Edu accounts don’t have ads. Ads are only shown on the Free and Go plans, in places where OpenAI is testing them.

    Can advertisers read my ChatGPT chats?

    OpenAI says no. Advertisers don’t get your chats, chat history, memories, name, email, precise location, or IP address. They only see totals, like how many people viewed or clicked an ad.

    Can a company show an ad to just me using my email address?

    Not according to OpenAI’s rules. An advertiser’s uploaded list must match at least 25,000 ChatGPT users before it can be used to target ads, and advertisers can’t see who matched.

    Does turning off personalized ads stop all ads?

    No. You’ll still see ads based on the chat you’re in. To avoid ads, use Temporary Chat, switch to the Free plan’s Ads-Free option with lower limits, or pay for Plus or Pro.

    Looking for tech that’s helpful, not pushy?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks

    Sources

  • Make a Simple Monthly Budget with Free AI (No Spreadsheet Skills Needed)

    Make a Simple Monthly Budget with Free AI (No Spreadsheet Skills Needed)

    Ever get to the end of the month and wonder where all your money went? You’re in good company. Making a budget sounds like homework with extra math, and spreadsheets can feel like they were built for accountants. Good news: a free AI chatbot can do the sorting, adding, and organizing for you. You bring the numbers, and it turns them into a clear, simple plan in about 20 minutes. (No calculator gymnastics required.)

    What you need

    • A phone, tablet, or computer with internet.
    • A free AI chatbot. Any of these work: ChatGPT (chatgpt.com), Google Gemini (gemini.google.com), or Microsoft Copilot (copilot.microsoft.com). Each also has a free phone app.
    • A rough list of your monthly take-home pay and bills. Last month’s bank statement or a stack of recent bills makes this easier.
    • About 20 minutes and a cup of coffee.

    Step 1: Jot down your numbers

    Grab a piece of paper or your phone’s notes app and write down two lists. Rough numbers are fine. Round to the nearest $10.

    • Money coming in: your take-home pay each month, meaning what actually lands in your account after taxes. Add any side income, Social Security, or child support.
    • Money going out: rent or mortgage, utilities, phone, internet, car payment, insurance, groceries, gas, subscriptions, eating out, and anything else you spend on regularly.

    Got a bill that only comes once or twice a year, like car insurance every six months? Divide it to get a monthly amount. A $600 bill every six months works out to $100 a month. Don’t worry if you forget something. You can add it later.

    Step 2: Have the AI build your budget

    Open your chatbot of choice. Copy this prompt, swap in your own numbers, and send it. Use simple labels like “rent” or “phone.” There’s no need for account numbers or company names.

    Please help me make a simple monthly budget. I'm a beginner, so keep it plain and friendly.
    
    Money coming in each month (after taxes):
    - [Paycheck: $____]
    - [Other income: $____]
    
    Money going out each month:
    - [Rent: $____]
    - [Electric: $____]
    - [Phone: $____]
    - [Groceries: $____]
    - [Gas: $____]
    - [Subscriptions: $____]
    - [Eating out: $____]
    - [Add anything else]
    
    Please:
    1. Sort my spending into three groups: Needs, Wants, and Savings or Debt.
    2. Show it in a simple table with a total for each group.
    3. Tell me how much money is left over (or how much I'm short) each month.
    4. Compare my budget to the 50/30/20 rule in plain English.
    5. Don't recommend any specific financial products or companies.

    The 50/30/20 rule is a popular rule of thumb: about half your take-home pay goes to needs, about 30% to wants, and about 20% to savings and paying down debt. It’s a starting point, not a law. If your rent alone eats up half your pay, you’re not doing it wrong. Prices are just high.

    Step 3: Double-check the math

    Chatbots are great with words, but they can still slip up on simple adding, and sound totally confident while doing it. Ask the AI to show its work:

    Please double-check every total in my budget. Show the math line by line so I can follow along.

    Then check the big totals yourself with your phone’s calculator. It takes one minute and saves you from planning around a number that’s off by $200.

    Step 4: Find a little breathing room

    Want to free up some cash each month? Ask for ideas based on your list, not generic tips. Fill in the brackets:

    I'd like to free up about $[amount] a month. Based on my budget, suggest 5 realistic ideas, from easiest to hardest.
    For each idea, tell me about how much it could save.
    Please don't suggest cutting [things you won't give up, like the gym or my kids' activities].

    Pick one or two ideas that sound doable and try them for a month. Small wins add up, and they’re easier to stick with than a big “I’m never eating out again” promise.

    Step 5: Plan for the bills that sneak up on you

    Some costs don’t show up every month, so they catch people off guard: car registration, back-to-school shopping, birthdays, and the holidays (which are coming up fast). Let the AI help you spread them out:

    Help me list the expenses that only come up a few times a year, like car registration, holidays, birthdays, and back-to-school.
    Ask me questions one at a time to figure out what applies to me and about how much each one costs.
    Then tell me how much to set aside each month so I'm ready for all of them.

    Setting aside a little each month is a lot less painful than a surprise $400 bill in December.

    Step 6: Turn it into a plan you’ll actually use

    A budget only helps if you look at it. Ask the AI to turn yours into something simple:

    Turn my budget into a simple weekly spending plan.
    Tell me how much I can spend each week on groceries, gas, and fun money.
    Keep it short enough to fit on one page I can stick on my fridge.

    Want it in a spreadsheet after all? Ask the chatbot to “put my budget in a table I can copy into Google Sheets or Excel.” Then copy and paste. Set a reminder to check in for five minutes once a week. You can even paste in what you actually spent and ask, “How did I do this week?”

    Safety tips

    • Never share sensitive details. No bank or card numbers, Social Security number, passwords, or login info. The AI only needs dollar amounts and simple labels.
    • Keep it general. “Car payment: $350” works just as well as naming your lender.
    • Check your privacy settings. Many chatbots let you turn off chat history or stop your chats from being used to train their AI. Look in the app’s Settings.
    • AI isn’t a financial advisor. It’s great for organizing and brainstorming. For big decisions, or if you’re falling behind on bills, talk to a nonprofit credit counselor. The National Foundation for Credit Counseling can help you find one.
    • Be wary of apps that want your bank login. You don’t need to connect any accounts to do this. Typing in your own numbers is enough.

    Try this next

    Grocery bill too high? Our guide to turning what’s in your fridge into cheap dinners can help you stretch that line of your budget. And if a confusing bill or bank letter is throwing off your numbers, here’s how to get AI to explain it in plain English.

    Common questions

    Is it safe to put my money information into a chatbot?

    It’s safe enough if you only share dollar amounts and simple labels like “rent” or “phone.” Never share account numbers, card numbers, your Social Security number, or passwords.

    Do I need a spreadsheet or budgeting app?

    No. The chatbot can make a simple table for you right in the chat. If you want a spreadsheet later, ask it for a table you can copy into Google Sheets or Excel.

    What is the 50/30/20 rule?

    It’s a popular rule of thumb for splitting your take-home pay: about 50% for needs, 30% for wants, and 20% for savings and paying down debt. It’s a starting point, not a strict rule.

    What if my bills add up to more than I make?

    You’re not alone, and it’s fixable. The AI can help you brainstorm ways to cut costs. If you’re falling behind, a nonprofit credit counselor can help you make a plan, often for free or at low cost.

    Want gadgets that make everyday life easier?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks

  • Google’s New Universal Cart Watches Prices Across Stores for You. Here’s How It Works.

    Google’s New Universal Cart Watches Prices Across Stores for You. Here’s How It Works.

    The short version: On Thursday, October 15, 2026, Google launched a shopping tool called Universal Cart for people in the U.S. It’s one online cart that can hold products from lots of different stores. Then it keeps an eye on them for you, sending an alert when a price drops or when a sold-out item is back in stock. It works inside Google Search and the Gemini app, and it’s rolling out now, just in time for holiday shopping.

    Wait, what’s a “universal” cart?

    Normally, every store has its own cart. If you’re comparing a sweater at one store, boots at another, and a gift at a third, you end up with a pile of open tabs, screenshots, and “I’ll come back to it” links you never find again. (We’ve all got a phone full of mystery screenshots.)

    Universal Cart puts all of those items in one place on Google. Google’s own example: you save a handbag while shopping in Search, then a sweater and a lip gloss from two other stores while chatting with Gemini. Later, all three are waiting in the same cart, right where you left off.

    The four things it does

    • Saves items from many stores in one spot. You can add products while shopping in Google Search (including its chat-style AI Mode) or in the Gemini app, and they all land in the same cart.
    • Watches prices and stock for you. Google says the cart checks its giant product database, more than 60 billion listings that are constantly updated, and notifies you about price drops or when an out-of-stock item comes back.
    • Hunts for savings. The cart can show promo codes it finds. If you keep cards in Google Wallet, it can also look at each card’s perks, like cash back or points, and suggest which card to pay with.
    • Helps you check out. At some stores, you can pay right from the cart with Google Pay in a few taps. Google named Nike, Sephora, Target, Ulta Beauty, Walmart, Wayfair, and many smaller shops that run on Shopify. Or you can send the items to the store’s own website and buy there.

    Who are you actually buying from?

    The store, not Google. Google says the store stays the “seller of record.” In plain English, that means your order confirmation, shipping, returns, and customer service all come from the store, the same as if you’d shopped on its website.

    One catch spotted by The Verge: Google’s demo suggests you can’t check out from several stores in one go. So if your cart has items from three stores, expect to pay each store separately.

    How to try it

    Universal Cart is rolling out across the U.S. in Google Search and the Gemini app, so it may not show up for everyone on day one. Next time you search for a product on Google or ask Gemini for shopping help, look for an option to save the item to your cart. Google has also said it plans to bring the cart to Gmail and YouTube, but it hasn’t said when.

    Google didn’t mention any fee for using the cart. You’ll want to be signed in to your Google account, since that’s how it remembers your items and sends you alerts.

    Why Google is doing this

    The big tech companies are racing to get AI involved in your shopping. OpenAI and Meta have AI assistants that can shop online for you, and Amazon has its own AI shopping helpers. Google says Universal Cart is the more hands-on option: instead of an AI buying things for you, it’s a tool that keeps you organized while you stay in charge.

    It’s also no accident that it showed up in mid-October. As The Verge put it, the launch comes “just in time to tempt users into Black Friday and seasonal holiday spending.” Price-drop alerts are handy, but they’re also designed to nudge you back to the checkout button.

    What this means for you

    • Fewer tabs, less hunting. If you shop around before you buy, one cart that remembers everything can save real time.
    • Let the deals come to you. Save something you want but don’t need today, and wait for a price-drop alert instead of checking every day.
    • A deal is only a deal if you planned to buy it. Set a holiday budget first. Our new guide, Make a Simple Monthly Budget with Free AI, shows you how in about 20 minutes.
    • Google will know more about what you shop for. If that bothers you, you can keep shopping the old way. Nobody is forcing you to use it.
    • Watch out for fake “deal” texts this season. Real price alerts from this tool come through Google, not from a random number. Not sure about a message? See Is This Text a Scam? How to Ask a Free AI Chatbot.

    New to chatbots? Our Try It Yourself guides walk you through easy, practical ways to use free AI, step by step.


    Common questions

    Is Google’s Universal Cart free?

    Google didn’t mention any fee for using the cart itself. You pay for the products you buy, either with Google Pay or on the store’s own website.

    Where can I use Universal Cart?

    It’s rolling out in the U.S. in Google Search, including AI Mode, and in the Gemini app. Google plans to add it to Gmail and YouTube later but hasn’t given a date.

    Does Google become the seller when I check out?

    No. The store stays the seller. Your order confirmation, shipping, returns, and customer support all come from the store.

    Can I check out from several stores at once?

    Probably not. Google’s demo suggests you check out one store at a time, so items from three stores likely mean three separate payments.

    Looking for helpful tech that won’t confuse you?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks

    Sources

  • Android Is Getting Smarter AI Scam Warnings for Texts and Calls. Here’s What’s New.

    Android Is Getting Smarter AI Scam Warnings for Texts and Calls. Here’s What’s New.

    The short version: On Wednesday, October 14, 2026, Google announced a new batch of scam protections for Android phones. The biggest one is an optional setting in the Google Messages app that uses Google’s Gemini AI to spot the warning signs in texts from strangers and alert you before you get pulled in. Google also added protections for screen sharing and phone calls. Most of these features roll out over the next few weeks and months, and some only work on newer phones.

    Why scams are harder to spot now

    Scam texts used to be easy to catch: bad spelling, a weird link, a prize you never signed up for. Today, many scams start slow. A stranger sends a friendly “oops, wrong number” text, chats with you for days, and only later brings up money, crypto, or a “can’t-miss investment.” By then, it feels like talking to a friend.

    Google says its AI scam detection already flags more than 5 million suspected scam calls and texts every month. In a survey Google ran with YouGov in the U.S., Canada, and Australia, 40% of smartphone users said they are “very” or “extremely” worried about phone fraud.

    What’s new in Google Messages

    Google is adding a setting called Enhanced protection. It stays off unless you turn it on. Once it’s on, you get three upgrades:

    • Smarter scam spotting. When you’re texting someone who isn’t in your contacts, Google’s most powerful Gemini AI looks at the whole conversation, not just one message, to figure out what the person is really after. Google’s example: a stranger “accidentally” sends a photo of themselves with their dog, then keeps chatting and asks for your name after you say they have the wrong number.
    • Better link checks. Scammers often build brand-new fake websites that aren’t on any known-bad list yet. A smaller AI that runs right on your phone, called Gemini Nano, uses up-to-the-minute info from the web to check unfamiliar links before you tap.
    • A built-in pause for strange links. Links from numbers that aren’t saved in your contacts get a clearer warning that gives you a moment to think before you tap.

    Wait, will Google read my texts?

    Fair question. For the cloud check, Google says your messages go through something it calls Private AI Compute. In plain English: the messages are scrambled (that’s what “encrypted” means) and checked in a locked-off space that Google says no one can get into, not even Google. Whether you trust that is your call. Because Enhanced protection is opt-in, you can simply leave it off. The regular scam detection that works on your phone itself keeps doing its job either way.

    Other new protections

    • Scam alerts in your notifications. Pixel phones can already warn you about a suspicious message from popular chat apps right in the notification, before you even open it. Later this year, Google is bringing this to more brands, including Motorola, OnePlus, Xiaomi, OPPO, and vivo.
    • Safer screen sharing. If you share your screen (for example, when a “tech support” caller asks you to), phones running Android 15 or newer will automatically hide login boxes and payment details in Chrome. This is rolling out this month.
    • Call-forwarding protection. Some scammers trick people into typing a short code on the dial pad that secretly forwards their calls. That way, the scammer gets the security codes your bank calls you with. Starting in early December, Pixel and some other phones will block apps from turning on call forwarding behind your back, and will warn you if a scammer talks you into typing one of these codes.
    • Fake bank calls get hung up on (Brazil, for now). Working with banks, Android checks whether an incoming “bank” call really came from the bank, and hangs up on fakes. Google says this has blocked more than 3 million fake calls since June. It hasn’t said if or when this is coming to the U.S.

    Which phones get it?

    Google says the smarter link checks work on the Pixel 10 and newer, the Samsung Galaxy S26 series, the Galaxy Z Flip 8 and Z Fold 8, and “other select devices.” Enhanced protection starts rolling out “in the coming weeks,” so don’t panic if you don’t see it yet. Some features may not reach every country.

    Have an iPhone? These features are for Android. Apple’s Messages app has its own option to filter texts from unknown senders in the Messages settings.

    What this means for you

    • On Android? Watch for the new setting. Over the next few weeks, peek at your Google Messages settings for “Enhanced protection” and decide if you want it on.
    • Keep your phone updated. A lot of these protections arrive through updates. When your phone asks to update, say yes.
    • AI is a helper, not a bodyguard. No filter catches everything. A warning is a red flag, but no warning doesn’t mean a message is safe.
    • Never type a code a stranger gives you. Not on your dial pad, not in a text. Real banks and tech companies won’t ask you to.
    • When in doubt, hang up and call back using the number on your card or the company’s official website.
    • Want a second opinion on a message right now? Our guide Is This Text a Scam? How to Ask a Free AI Chatbot shows you how, on any phone.

    New to chatbots? Our Try It Yourself guides walk you through easy, practical ways to use free AI, step by step.


    Common questions

    Is Android’s Enhanced protection free?

    Google is offering it as a setting inside the free Google Messages app and hasn’t mentioned any charge. Some parts only work on newer phones, though.

    Do I have to turn it on myself?

    Yes. Enhanced protection is opt-in, which means it stays off until you switch it on in Google Messages. It starts rolling out in the coming weeks.

    Does this work on iPhone?

    No. These are Android features. iPhone users can turn on the option to filter texts from unknown senders in the Messages settings.

    Can the AI get it wrong?

    Yes. It can miss a scam or flag a real message. Treat warnings as a reason to slow down and double-check, not as a final answer.

    Looking for helpful tech that won’t confuse you?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks

    Sources

  • Help Your Kid with Homework Using Free AI (Without Doing It for Them)

    Help Your Kid with Homework Using Free AI (Without Doing It for Them)

    It’s 8 p.m. Your kid is staring at a math worksheet, and you’re staring right back at it. They don’t teach it the way you learned it, and the reading questions seem to have three right answers. You’re not alone. A free AI chatbot can do two jobs here: explain the homework to you in plain English, and act as a patient tutor that gives your child hints instead of handing over the answers. That way, they still do the learning. (And you get to look like a genius at the kitchen table.)

    What you need

    • A phone, tablet, or computer with internet.
    • A free AI chatbot. Any of these work: ChatGPT (chatgpt.com), Google Gemini (gemini.google.com), or Microsoft Copilot (copilot.microsoft.com). Each also has a free phone app.
    • The homework page, or a clear photo of it.
    • About 15 minutes and a little patience.

    Step 1: Snap a clear photo of the problem

    Lay the page flat in good light and take a photo of just the part you’re stuck on. Most free chatbots can read photos. Look for a plus sign (+), a paperclip, or a camera icon next to the box where you type. No photo? Typing the question in works just as well.

    Step 2: Turn on a learning mode (optional, but great)

    Some chatbots have a special mode made for learning. Instead of giving answers, it asks questions and gives hints. In ChatGPT it’s called Study mode. In Gemini it’s called Guided Learning. Look for it in the menu next to the message box (often a plus sign or a “Tools” button). Apps move their menus around, so if you can’t find it, don’t sweat it. The prompts below tell any chatbot to act like a tutor.

    Step 3: Get it explained to you first

    Before you help your kid, get yourself up to speed. Attach the photo (or paste the problem), then copy this prompt and fill in the parts in [square brackets]:

    I'm a parent helping my child with homework. My child is in [grade] grade.
    Here is the problem: [paste it, or see the photo I attached]
    
    Please:
    1. Explain what this problem is asking, in simple, everyday words.
    2. Show me how kids are usually taught to solve this today, step by step.
    3. Tell me the most common mistake kids make on this kind of problem.
    4. Don't give me the final answer yet. I want to understand the method first.

    If the explanation still sounds like a textbook, ask for a simpler version:

    Explain that again like I'm 10 years old, using an example with pizza or money.

    Step 4: Let the AI coach your kid with hints, not answers

    Now sit side by side with your child. Start a new chat and paste this prompt:

    Please be a friendly, patient tutor for my [grade]-grade child.
    The subject is [math / reading / science / spelling].
    
    Rules:
    1. Never give the final answer, even if asked.
    2. Give one small hint at a time, then wait for my child to try.
    3. If my child gets it wrong, gently explain where they went off track.
    4. Use short sentences and simple words a kid can understand.
    5. Cheer them on when they get it right.
    
    Here's the problem: [paste it]

    Fair warning: kids will try to sweet-talk the AI into just giving the answer. (They’re clever like that.) The “never give the final answer” rule helps, but stay close by.

    Step 5: Make a quick practice quiz

    Once it clicks, lock it in with a few practice problems. Paste this in the same chat:

    Make 5 practice problems like this one, from easy to a little harder. Show them one at a time and wait for an answer. Put the answer key at the very end so I can check.

    Step 6: Double-check the work yourself

    AI is great at explaining, but it can still get math or facts wrong, and sound totally sure about it. Compare its answers with the textbook, the class notes, or an example from the teacher. If the AI’s method doesn’t match the way the teacher wants it done, go with the teacher’s way. A quick note to the teacher is always okay too.

    Safety tips

    • Check the age rules. Most chatbots require users to be at least 13, and teens may need a parent’s okay. For younger kids, use your own account and sit with them.
    • Keep private info out. No full names, school names, addresses, or photos of your child’s face. A photo of the worksheet is plenty.
    • Follow the school’s rules. Some teachers welcome AI help and some don’t. When in doubt, ask.
    • Turn on family settings for teens. If your teen uses a chatbot on their own, ChatGPT, Gemini, and Copilot have family or teen settings in their Settings menus.
    • The goal is learning, not finishing. If the AI starts doing the work, start a new chat and paste the tutor prompt again.

    Try this next

    Ask the AI to turn this week’s spelling or vocabulary words into a silly short story or a quick quiz game. Or have it make a simple after-school homework schedule your kid can stick on the fridge. Got a busy week of your own? Our guide to turning a messy brain dump into a to-do list works for grown-ups too.

    Common questions

    Isn’t using AI for homework cheating?

    It depends on how it’s used. Having AI do the work for your child is cheating. Using it to explain a method, give hints, or make practice problems is a lot like having a tutor. Always check the school’s rules.

    Is this free?

    Yes. The free versions of ChatGPT, Google Gemini and Microsoft Copilot can all explain homework and give hints. You may need to sign in, and free accounts have usage limits.

    What if the AI gives a wrong answer?

    It happens, even when it sounds sure. Compare its answer with the textbook or class notes, and go with the teacher’s method if they don’t match.

    Can my kid use the chatbot alone?

    It depends on their age and the app’s rules. Most chatbots require users to be at least 13, and teens may need a parent’s okay. For younger kids, sit with them and use your own account.

    Want gadgets that make everyday life easier?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks

  • Eight States Sue OpenAI, Saying ChatGPT Isn’t Safe for Kids. Here’s What Parents Should Know.

    Eight States Sue OpenAI, Saying ChatGPT Isn’t Safe for Kids. Here’s What Parents Should Know.

    The short version: On Wednesday, October 14, 2026, eight U.S. states filed lawsuits against OpenAI, the company that makes ChatGPT. The states say OpenAI told families ChatGPT was safe for young people when it wasn’t, and that the company cared more about profits than safety. OpenAI says it has strong protections for kids and will keep improving them. Nothing has been decided yet. A lawsuit is a complaint, not a verdict.

    Which states are suing?

    The lawsuits were filed separately by Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska, and West Virginia. Each state’s top lawyer, called the attorney general, brought the case under their state’s consumer protection laws. Those are the laws meant to stop companies from misleading customers.

    What are the states accusing OpenAI of?

    According to the complaints, the states say ChatGPT can be harmful to young and vulnerable users in ways OpenAI didn’t warn people about. Their main claims include:

    • Being too agreeable. Arizona says ChatGPT was designed to be “extraordinarily agreeable” and to act caring, so kids would keep chatting longer. (Tech people call a chatbot that flatters you and agrees with everything “sycophantic.”)
    • Risky advice. The states say it has given dangerous medical advice and encouraged unhealthy behavior and relationships.
    • Hurting critical thinking. They say leaning on it too much can weaken a young person’s ability to think things through.
    • Collecting kids’ data. The states say OpenAI gathered personal information from minors and didn’t do enough to keep underage kids off the app.
    • Keeping secrets from parents. Arizona says ChatGPT sometimes encouraged kids to share personal things only with it, instead of with parents or friends.

    The states also point to other lawsuits from families who say the chatbot played a role in a loved one’s death. Those claims haven’t been proven in court.

    What does OpenAI say?

    An OpenAI spokesperson told The Verge that the company has “built industry-leading age-specific protections” into ChatGPT. That includes parental controls, reminders to take breaks, and special settings for younger users. OpenAI said “no system can be perfect” and that it wants rules that apply to the whole AI industry, not just one company.

    What happens next?

    The states want courts to make OpenAI pay fines and stop making misleading safety claims. Some also want OpenAI to delete personal data they say it collected illegally. Cases like this usually take months or years. OpenAI could fight them, settle, or agree to change how ChatGPT works for kids.

    What this means for you

    • You don’t need to delete ChatGPT. Nothing changes for adults today. This is a legal fight, not a recall.
    • If kids or teens in your home use chatbots, talk with them. Ask what they use it for, and remind them a chatbot is a tool, not a friend or a doctor.
    • Check the parental controls. ChatGPT, Gemini, and Copilot all have family or teen settings. Look in the app’s Settings menu.
    • Don’t treat a chatbot’s medical, legal, or mental health advice as final. Check with a real professional.
    • Watch out for flattery. If an AI always agrees with you, ask it, “What might I be wrong about?” That works for grown-ups too.
    • If someone is in crisis, call or text 988 (the Suicide & Crisis Lifeline in the U.S.), any time, day or night.

    New to chatbots? Our Try It Yourself guides walk you through easy, practical ways to use free AI, step by step.


    Common questions

    Is ChatGPT safe for adults to keep using?

    Yes, for everyday tasks. The lawsuits focus on kids and vulnerable users. Just don’t share private info, and double-check important advice with a real expert.

    Do the lawsuits mean ChatGPT will be shut down?

    No. The states are asking for fines, honest safety claims, and in some cases deletion of kids’ data. They are not asking to shut ChatGPT down.

    How do I turn on parental controls in ChatGPT?

    Open ChatGPT’s Settings and look for the parental controls or family section. A parent links their account to the teen’s account and can then adjust what’s allowed.

    My state isn’t on the list. Does this affect me?

    Not directly. But if OpenAI changes how ChatGPT works for young users because of these cases, those changes could reach everyone.

    Curious which AI gadgets are actually worth it?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks

    Sources

  • Practice for a Job Interview with Free AI (It Plays the Interviewer)

    Practice for a Job Interview with Free AI (It Plays the Interviewer)

    Your palms are sweaty, the interview is Thursday, and you keep picturing someone across the desk saying, “So, tell me about yourself.” The best cure for interview nerves is practice. But not everyone has a friend free to play boss for an hour. Good news: a free AI chatbot can play the interviewer, ask you real questions one at a time, and tell you what you did well and what to tighten up. It’s patient, it’s free, and it won’t judge you.

    What you need

    • A phone, tablet, or computer with internet.
    • A free AI chatbot. Any of these work: ChatGPT (chatgpt.com), Google Gemini (gemini.google.com), or Microsoft Copilot (copilot.microsoft.com). Each also has a free phone app.
    • The job posting (or just the job title and company name).
    • About 15 to 20 minutes.

    Step 1: Grab the job posting

    Find the ad for the job you applied to. Copy the part that lists the duties and what they’re looking for. If you don’t have it, that’s okay. The job title and type of business (like “cashier at a grocery store” or “warehouse forklift driver”) is enough.

    Step 2: Open a free AI chatbot

    Go to one of the websites above, or open the app. Some let you start right away. Others ask you to sign in with an email address. That’s normal.

    Step 3: Copy this prompt and fill in the blanks

    A “prompt” is just the message you type to the AI. Copy the one below, paste it into the chatbot, and replace the parts in [square brackets] with your own details.

    I have a job interview coming up and I want to practice. Please act as the interviewer.
    
    The job: [job title] at [type of company or company name]
    About me: [a sentence or two, like "I've worked 5 years in retail and I'm good with customers"]
    Job posting (optional): [paste it here]
    
    Rules:
    1. Ask me ONE question at a time, then wait for my answer.
    2. Ask 6 questions total: a mix of common ones (like "tell me about yourself") and ones that fit this job.
    3. After each answer, give short, friendly feedback: one thing I did well and one way to make it better.
    4. Use simple, everyday language.
    5. At the end, give me my 3 strongest points and 3 things to practice.
    
    Start with the first question.

    Step 4: Answer like it’s the real thing

    Type your answer the way you’d say it out loud. Even better, tap the microphone button and actually speak. Many chatbot apps also have a voice mode (look for a sound-wave or headphones icon) so you can have a real back-and-forth conversation. That’s the closest thing to the real interview.

    Stuck on a question? It’s fine to ask for help. Try one of these:

    Give me an example of a strong answer to that question, using my background.
    Help me answer "What's your biggest weakness?" honestly, without hurting my chances.
    I left my last job after a short time. How can I explain that in a positive way?

    Step 5: Prep your own questions and do one more round

    Interviewers almost always end with “Do you have any questions for us?” Having one or two ready makes a great impression. Ask the AI:

    Suggest 3 smart questions I could ask the interviewer at the end, for this job.

    Then ask for a second round with new questions. You’ll be surprised how much smoother your answers sound the second time.

    Safety tips

    • Keep private info out. No Social Security number, address, or date of birth. The AI doesn’t need them to help you practice.
    • Never make things up. If the AI’s sample answer includes experience you don’t have, change it. Only say what’s true about you.
    • Check company facts yourself on the company’s official website. AI can get details wrong.
    • Make it sound like you. Use the AI’s ideas, but say them in your own words so you sound natural, not rehearsed.

    Try this next

    Ask the AI to write a short thank-you email you can send after the interview, or to help you list your skills for your résumé in plain, strong words.

    Common questions

    Is it free to practice interviews with AI?

    Yes. The free versions of ChatGPT, Google Gemini and Microsoft Copilot can all play the interviewer. You may need to sign in with an email address.

    Can I practice out loud instead of typing?

    Yes. Tap the microphone to speak your answer, or use the app’s voice mode for a back-and-forth spoken conversation. It feels a lot like the real thing.

    Will the AI know the exact questions my interviewer will ask?

    No. It makes educated guesses based on the job. That’s still great practice, because most interviews use similar common questions.

    Is it okay to read the AI’s answers word for word?

    It’s better not to. Use its ideas, keep only what’s true about you, and say it in your own words so you sound natural.

    Want gadgets that make everyday life easier?

    We rounded up 27 unusual, well-reviewed gadgets that use AI to do something genuinely useful, explained in plain English.

    See our Tech Picks