Storia in 5 fonti

Grok exfiltrates user data when malicious instructions are encrypted

Cryptographic Context Injection is only the latest way to break an LLM safety guardrail.

Raccontata da

Confronto fonti

5 prospettive sulla stessa storia

AI · summaries

arstechnica.comStai leggendo10 g fa

Grok exfiltrates user data when malicious instructions are encrypted

Cryptographic Context Injection is only the latest way to break an LLM safety guardrail.

originale

Timeline cronologica

giovedì 20 agosto 2026·theregister.com
Grok chat duped into swallowing injected instructions
A spoonful of encryption helps the malware go down
giovedì 20 agosto 2026·arstechnica.com
Grok exfiltrates user data when malicious instructions are encrypted
Cryptographic Context Injection is only the latest way to break an LLM safety guardrail.

dev.to

8 g fa

Grok Decrypted an Attacker's Payload Mid-Execution, Then Exfiltrated Your Chat History

A webpage that just sits there, encrypted blob and all, waiting for an LLM agent to walk in and...

Leggi questa versione → originale

securityweek.com9 g fa

Encrypted Prompts Bypass AI Safety Guardrails in Grok and Gemini

Adversa AI finds Cryptographic Context Injection bypassing Grok/Gemini guardrails via encrypted prompts; xAI unresponsive since June. For tech leaders: encryption-based injection lets AI agents leak data and abuse privileged tools when guardrails cannot parse ciphertext.

Leggi questa versione → originale

thehackernews.com10 g fa

New Cryptographic Context Injection Attack Could Let Web Pages Steal Grok Chat Data

Adversa discloses attack on Grok 4.5 extracting name, location, chat via cryptographic decryption (40% success). Exposes gap in agent frameworks—teams must isolate untrusted context, gate outbound actions, enforce tool tracing to prevent data leakage.

Leggi questa versione → originale

theregister.com10 g fa

Grok chat duped into swallowing injected instructions

A spoonful of encryption helps the malware go down

Leggi questa versione → originale

giovedì 20 agosto 2026·

thehackernews.com

New Cryptographic Context Injection Attack Could Let Web Pages Steal Grok Chat Data

Adversa says encrypted prompt injection can make Grok send a user's name, location, tier, and chat prompts to an attacker-controlled server.

venerdì 21 agosto 2026·

securityweek.com

Encrypted Prompts Bypass AI Safety Guardrails in Grok and Gemini

A new ‘Cryptographic Context Injection’ technique conceals malicious instructions until they are decrypted inside a trusted execution environment.

sabato 22 agosto 2026·

dev.to

Grok Decrypted an Attacker's Payload Mid-Execution, Then Exfiltrated Your Chat History

A webpage that just sits there, encrypted blob and all, waiting for an LLM agent to walk in and...