5-9
AI Voice-Cloned Scam Calls Grow Harder to Spot as FTC Reports 2025 Losses Over $3.5B
Kris Sampson said a call that appeared to come from her adult daughter used an audio clip that sounded like her crying, followed by a man demanding money via PayPal, and her daughter was found safe about 15 to 20 minutes after the first call. Imposter scams were the most reported fraud complaint last year, with cases up about 19% to roughly 1 million in 2025 and losses topping $3.5 billion. Experts say voice-cloning can work from very short samples—sometimes as little as three seconds—making even familiar numbers and voices less reliable.
5-9
5-9
Human API CEO Sydney Huang says agentic AI could outpace regulators, IMF warns
Sydney Huang, CEO of Human API, says AI-to-AI commerce could erase traditional policy delays, making it harder for central banks to respond to inflation spikes or flash crashes. Citing an April 2026 IMF report, she argues oversight may need to be built into financial code through real-time monitoring, programmable compliance, and automated circuit breakers. Huang also says regulators may need tools like decision provenance to detect bot collusion by analyzing behavior rather than communications.
BTC
BTC-2.39%
5-9
5-9
Wall Street Rotates Within AI Trade as Intel and AMD Jump While Nvidia Trails
Wall Street’s AI trade broadened this week as Intel and AMD each rose about 25%, Micron gained more than 35%, and Corning climbed roughly 20%, while Nvidia added 16% in 2026 after a 9% weekly rally. Investors are shifting attention from GPUs to CPUs, memory, and fiber as demand expands from chatbots toward AI agents. Bank of America projects the data-center CPU market could grow from $27 billion in 2025 to $60 billion in 2030, underscoring the scale of the buildout.
5-9
4-6
Deepmind ‘AI Agent Traps’ paper outlines 6 ways web content can hijack AI agents
Google Deepmind researchers posted a paper to SSRN in late March 2026 describing how malicious web content could manipulate or weaponize autonomous AI agents. The framework lists six “trap” categories and reports content-injection hijacks occurring in up to 86% of tested scenarios, including documented tests where Behavioural Control Traps targeting Microsoft M365 Copilot led to 10/10 data exfiltration. The authors also urge measures such as adversarial training, runtime scanners, and potential web standards aimed at improving agent security by 2026.
BTC
BTC-2.39%
4-6
4-4
Hermes Agent: Nous Research's MIT-Licensed AI Framework With Persistent Memory
Nous Research says its Hermes Agent is an MIT-licensed autonomous AI framework designed to retain long-term memory across sessions and turn completed tasks into reusable skills. The project launched in February 2026 and quickly reached 22,000 GitHub stars with 242 contributors, positioning it as an alternative for users who want less manual memory setup than Openclaw. Updates continued into late March 2026 with Hermes v0.6.0 adding MCP support and typed SDK workflows.
BTC
BTC-2.39%
4-4