Spoofed AI sidebars can trick Atlas, Comet users into dangerous actions
OpenAI's Atlas and Perplexity's Comet browsers are vulnerable to AI sidebar spoofing attacks that mislead users into following fake AI-generated instructions. [...]
Spoofing impersonates trusted users, devices, or services to bypass trust and cause fraud; verify identities with strong authentication and signed messages.
Search across headline titles and summaries.
Background for this topic.
Spoofing is the forging of an identity or origin so a message, connection, website, or phone call appears to come from a trusted source. Attackers may impersonate an executive by email, imitate a domain, falsify caller ID, or place packets with a forged source IP address. The goal can be to induce payment or disclosure, deliver malware, bypass trust checks, or obscure the source of traffic. IP spoofing usually prevents a reply from reaching the attacker, but can support reflection attacks; it is not by itself proof of access to the claimed system.
Mitigation depends on the channel. For email, configure and enforce SPF, DKIM, and DMARC, while treating display names and caller ID as untrusted signals. Use phishing-resistant authentication, verify sensitive requests through an independent channel, and validate domains and certificates before users enter credentials. At network boundaries, apply ingress and egress source-address filtering, monitor anomalous traffic, and design services not to trust a claimed address alone.
OpenAI's Atlas and Perplexity's Comet browsers are vulnerable to AI sidebar spoofing attacks that mislead users into following fake AI-generated instructions. [...]
Security pros explore whether infection-spoofing code can immunize Windows systems against attack Feature What's better, prevention or cure? For a long time the global cybersecurity industry has operated by reacting to attacks and computer viruses. But given that ransomware has continued to escalate, more proactive action is needed.…