AI News Brief for Thursday, May 21, 2026

  • WhatsApp's New Incognito Mode Keeps Your AI Conversations Truly Private.
    A new Incognito Chat mode on WhatsApp processes your AI questions in a secure environment even Meta cannot access, and erases them when you're done.

  • AI Scams are Surging. Here's How to Protect Yourself.
    The FBI logged $893 million in AI fraud losses in 2025, with adults 50+ among the prime targets. Here are the scams to watch for and the steps to take if something feels off.

  • New Bipartisan Federal Bill Would Stop AI from Charging You More at the Grocery Store.
    New bipartisan legislation targets AI algorithms that quietly charge you more than your neighbor for the same groceries, based on what they know about your spending habits.

WhatsApp's New Incognito Mode Keeps Your AI Conversations Truly Private

Privacy & Data Security
Source: Meta / Facebook
https://about.fb.com/news/2026/05/incognito-chat-whatsapp-meta-ai/

In a nutshell

Meta launched Incognito Chat on WhatsApp and the Meta AI app, the first major AI product where conversations leave no server log, not even accessible to Meta. Inference runs in a Trusted Execution Environment, conversations disappear on exit, and no data is stored or used to train future models. Intended for sensitive topics such as health questions, financial decisions, and career advice. Note: Source has a commercial interest in this topic.

Note: Use this app with caution and read the privacy policy before disclosing any sensitive personal information. Also, verify consequential medical, legal or financial advice with a professional before acting on it.

Why it's important

Adults 50+ who use WhatsApp and worry about AI tracking what they type about health conditions, finances, or family matters now have a verified way to ask without creating a permanent record.

Key takeaways

  • Incognito Chat runs inside a Trusted Execution Environment, meaning not even Meta employees can read your conversations, which goes a step beyond other "private" or incognito browser modes that still log data on the company's end.

  • Conversations disappear by default when you exit the chat, and no data from your session is stored or used to train future AI models, so questions about sensitive topics leave no trace.

  • The feature is designed specifically for personal and sensitive subjects, such as health questions, financial decisions, and career advice, where people may want to seek information without a permanent record.

  • Incognito Chat is rolling out gradually on both WhatsApp and the Meta AI app over the coming months, so not all users will see it right away.

  • Meta also plans to introduce a "Sidechat" feature using the same Private Processing technology, allowing private AI assistance within an active WhatsApp conversation without interrupting the main chat.

Your thoughts?

Have you ever held back from asking an AI a personal health or financial question because you were unsure where your words would end up? Would a truly private mode change how you use these tools?

AI Scams are Surging. Here's How to Protect Yourself.

Privacy & Data Security
Source: Forbes
https://www.forbes.com/sites/technology/article/ai-generated-scams

In a nutshell

This article is basically a refresher on the most common tactics (voice cloning, romance fraud, fake job interviews via deepfake video, PDF-based credential theft, and fake e-commerce sites). It provides concrete protective steps including independent verification, immediate fund recall, and multi-factor authentication. Note: Several quoted experts represent cybersecurity vendors with a commercial interest in this topic.

Why it's important

Adults 50+ are specifically identified as prime targets for AI confidence scams and investment fraud, where cloned voices mimic family members or bank representatives to steal money. The article gives actionable steps: verify callers through a separate channel, freeze credit immediately, and turn on call screening, all directly relevant to protecting retirement savings.

Key takeaways

  • Seniors and people living alone are prime targets for AI confidence and investment scams, where voice-cloning technology built from just seconds of social media audio can make a scammer sound exactly like a distressed family member calling for emergency help.

  • If you receive an unexpected urgent call from a loved one, hang up and call them back on a number you already have, or ask a question only they would know; deliberate verification is the single most effective defense.

  • PDF-based scams use free online document editors that silently log every personal detail you type, then use that real information to make follow-up phishing calls or emails nearly impossible to recognize as fake.

  • Fake e-commerce sites now come with polished product listings and responsive chatbots; check URLs carefully before purchasing, and treat any site you found through an unsolicited link with extra caution.

  • If you suspect you have been scammed, call your bank immediately to request a wire recall (you may have only hours), freeze your credit with all three bureaus, change passwords on every account involved, and enable multi-factor authentication using an authenticator app rather than SMS.

Your thoughts?

Do you feel confident that you can spot a scam when you see one?

New Bipartisan Federal Bill Would Stop AI from Charging You More at the Grocery Store.

Policy & Regulation
Source: Rep. Josh Gottheimer
https://gottheimer.house.gov/posts/release-gottheimer-announces-new-bipartisan-legislation-to-ban-ai-surveillance-pricing-at-grocery-stores-delivery-services

In a nutshell

Congressman Gottheimer (D-NJ) and Congressman Lawler (R-NY) introduced the No Rigged Grocery Prices Act, legislation to prevent stores and delivery apps from using AI-powered personal data to charge different customers different prices for identical items. New Jersey shoppers found the same grocery item priced up to 23% higher for some customers. The bill would require price-change disclosures and ban AI chatbots from substituting items without explicit consent.

Why it's important

Older adults on fixed incomes are particularly vulnerable to AI-driven price manipulation at grocery stores. This bill, announced today, directly addresses algorithms that use your personal shopping data (which may also correlate with age) to charge you more than the person standing next to you.

Key takeaways

  • Instacart was caught charging some customers up to 23% more for the exact same item at the same store and the same moment, costing a family of four roughly $1,200 more per year, and triggering FTC and state attorney general investigations as well as a separate $60 million fine for other deceptive practices.

  • The No Rigged Grocery Prices Act, introduced May 18, 2026, would prohibit grocery stores and delivery apps from using AI-driven personal data to set individualized, higher prices for identical items.

  • The bill requires stores to make clear disclosures whenever prices legitimately change (for example, items sold by weight or supply shortages), so shoppers are never surprised by unexplained price differences.

  • AI chatbots used in grocery delivery apps could be prohibited from substituting ordered items with alternatives unless the customer gives explicit consent, closing a loophole that has frustrated many shoppers.

  • Discounts, coupons, loyalty programs, and promotional pricing would all still be permitted under the legislation, so genuine savings tools remain available.

  • Maryland has already passed the first state law banning personal-data-driven price increases, and New Jersey is advancing similar legislation, reflecting growing bipartisan momentum to address this issue.

Your thoughts?

Have you noticed prices for your regular grocery items shifting unexpectedly, especially when ordering online or through a delivery app?

This brief was prepared by Martin Danner using Claude Cowork as a research assistant.