
Anthropic just confirmed everyone's worst fear
Wes Roth16 August 2026Watch on YouTube
Description
Build your own real-time voice agent with ElevenAgents and get 10,000 free credits to play around with: ➡️ https://try.elevenlabs.io/wes I put three ElevenAgents through real customer-service stress tests: an ecommerce store, a smart-home help desk, and an internet provider. I tested whether they could follow real policies, respond naturally under pressure, and resist prompt-injection attempts. ElevenAgents can connect with business tools, support conversations across 70+ languages, and use Expressive Mode to respond with more natural tone and emotion. Build an agent for your business, load in your actual policies, and call it as your own worst customer. That’s the benchmark that matters. ______________________________________________ My Links 🔗 ➡️ Twitter: https://x.com/WesRoth ➡️ AI Newsletter: https://natural20.beehiiv.com/subscribe Want to work with me? Brand, sponsorship & business inquiries: wesroth@smoothmedia.co Check out my AI Podcast where me and Dylan interview AI experts: https://www.youtube.com/playlist?list=PLb1th0f6y4XSKLYenSVDUXFjSHsZTTfhk ______________________________________________ 00:00 Anthropic's Research 03:48 ElevenLabs (sponsor) 08:42 AI turf wars 15:35 Mythos Strikes First #ai #openai #llm
What you'll learn
- ElevenAgents voice AI systems are tested in realistic customer-service environments, including e-commerce, smart-home support, and internet providers
- Voice agents can implement and enforce business policies, but must undergo stress tests to validate their reliability
- ElevenAgents support communication in 70+ languages and can use Expressive Mode to deliver more natural and emotional tone
- Voice agents can be integrated with business tools, but must also be tested against prompt-injection attacks