How McAfee's AI fights tech scams and protects users

A poorly-architected fraud system doesn't just leak money, it actively drives away good customers ➡️ when you falsely decline a legitimate user, there's a 27% chance they will never return. The fix isn't just "better rules" but rather a complete architectural shift. A modern, 3-layer, pre-authorization stack is the only way to scale: 1️⃣ The front line: bot detection & device fingerprinting 2️⃣ The transaction: velocity checks & throttling 3️⃣ The brain: AI/ML models & custom rule engines This is the model we use. The result? Our average chargeback rate is 0.3%, less than half the 0.8% industry average. Keen to know more about our fraud detection? Link in comments👇

The most consequential finding from this year's Global Bot Security Report isn’t just that AI traffic is up, but where that traffic is going. The common assumption is that LLM crawlers skim public pages for training data. Our data shows a different reality: AI traffic is increasingly interacting with the same high-value flows that drive revenue and risk. 💡 64% of AI traffic visits forms 💡 23% visits login pages 💡 Measurable volumes also hit carts, payment & account creation pages In practical terms, this is where fraud manifests: fake lead inflation via forms, ATO & credential stuffing at login, and checkout automation that can turn into card testing or scalping. Jérôme Segura explains more in Dark Reading... link in comments!

Is your business prepared for when AI goes wrong? AI is already changing how we work, think, and do business, but what happens when the tech that powers your future becomes your biggest risk? From fake data and biased algorithms to privacy breaches and black-box decisions, AI could be the next corporate crisis waiting to happen. 👉 Read the full analysis in this week’s Vellum here: https://bit.ly/3JAWfAX #KeepItVellum

AI is reshaping how businesses operate — and the numbers tell the story. Over half of business owners now use AI for cybersecurity and fraud prevention. 97% believe ChatGPT will help their business. A third plan to use it for website content — and nearly half for content in other languages. 46% already use AI for internal communications. Still, 40% worry about over-dependence on technology. Yet nearly two-thirds believe AI will improve customer relationships. AI isn’t just a trend — it’s becoming the backbone of modern business. How is your company using it? #AIinBusiness #WorkSmarter #ArtificialIntelligence Full article: https://lnkd.in/gSqgwFMW

In my opinion, Stephen H. talks a lot of sense about AI. This post is especially worthy of your attention. My two cents’ worth: AI companies are not driven by a desire to improve our lives. They will steal anything and lie to anyone in order to improve their bank balances, regardless of the consequences for anyone else. The last thing we should do is let them control and monitor our access to the online services we rely on to live our lives.

Your Brand Can Be Faked in Minutes. As AI becomes more powerful and sophisticated, creating fake content is now easier than detecting it. Deepfakes spread quickly and are hard to verify at scale. The consequences are real: lost funds, security breaches, damaged reputations and brands, and even challenges to sovereignty across enterprise, consumer, and government ecosystems. They also erode privacy and democratic trust. With usage surging, it is critical to deploy reliable tools and techniques that can detect deepfake information. Why this matters 1. Finance: payment fraud, wire scams, market manipulation 2. Security: account takeover, MFA workarounds, insider lures 3. Reputation and brand: executive impersonation, fake announcements, non-consensual content 4. Civic space: voter deception, information ops, media tampering Make this your baseline 1. Verify requests for money or access with a second channel. 2. Flag unverified audio, video, and images as untrusted until proven otherwise. 3. Train staff on the playbook: pause, verify, escalate. 4. Pilot detectors and provenance checks. Track precision, recall, and false positives. 5. Monitor for misuse of your brand and executive likenesses. Tools worth testing 1. Sensity AI: https://sensity.ai/ 2. DeepSecure-AI (open source): https://lnkd.in/eGXrHtWs 3. Intel Trusted Media detection research: https://lnkd.in/eyDyPk-C 4. Pindrop (voice security): https://www.pindrop.com/ 5. AI Voice Detector: https://lnkd.in/enJvSiYf 6. Hugging Face Space: Generalizable Deepfake Detection (NPR-CVPR 2024): https://lnkd.in/eZ5cJhwD #AI #Deepfakes #Security #FraudPrevention #BrandProtection #TrustAndSafety

Reeler’s favorite catch of the day? Your credentials. 🪝 He hides behind fake login pages and convincing sender names—and once you bite, it’s too late. Keep your guard up this #CybersecurityAwarenessMonth and learn from Abnormal AI how to recognize his scams before they strike. 👉 https://lnkd.in/eM3fBB_D

There needs to be the concept of trusted-source or trusted-device – that the AI will accept instructions only from certain terminals or sources. "David" logged in with corporate credentials (and 2FA/3FA) via an internal IP address (including corporate VPN IP addresses), sure. Random user of web page, NFW. The AI hack that convinced a chatbot to sell a $76,000 car for $1 https://lnkd.in/erJTcY5c  

That document might not be real - and you’d never know it. Document fraud is up 300% year-over-year. AI’s flipped the script. It’s made it easier than ever to create great customer experiences - but it’s also made it way easier to fake documents that look 100% legit. In this short clip from our chat with Resistant AI - Fraud Detection, I share what’s driving that 300% spike and why fast, automated processes still need a strong layer of verification. Because speed without trust? That’s a short-term win and a long-term risk. Watch the 1-min clip below and it might make you rethink how you’re handling document reviews. Full 5-minute recap link in the first comment.

Well, black hat hackers (and porn site marketeers) are the first to take full advantage of any technology. AI Agents wouldn't be an exception. "Exclusive | Chinese Hackers Used Anthropic’s AI to Automate Cyberattacks"! https://buff.ly/rR6BhTU