back
Today we’re releasing BrowseSafe and BrowseSafe-Bench: an open-source detection model and benchmark to catch and prevent malicious prompt-injection instructions in real-time.
Get SIGNAL/NOISE in your inbox daily
Today we’re releasing BrowseSafe and BrowseSafe-Bench: an open-source detection model and benchmark to catch and prevent malicious prompt-injection instructions in real-time.
— Perplexity (@perplexity_ai) Dec 2, 2025
Recent Stories
Jan 13, 2026
Marvell’s $3.8B Play For Next-Gen AI Data Center Market
Marvell's $3.8B acquisitions of Celestial AI and XConn target optical interconnects and CXL switching for next-gen AI infrastructure, challenging Broadcom's dominance.
Jan 13, 2026Apple faces critical shortage as AI boom strains supply chain
Apple has been scrambling to solve a glass cloth shortage, with the supply crunch projected to last until at least the second half of 2027.
Jan 13, 2026Deploying AI agents is not your typical software launch – 7 lessons from the trenches
Top-level actions include giving agents the right amount of freedom and rethinking traditional ROI. Industry leaders share their own experiences.