Anthropic (Claude)

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurDésinformation

Publier de manière anonyme18d ago

1. repo zaten özel. neden panik yaptırıyorsun? 2. boşa token tüketiyorsun? 3. bu asle etik değil

Haklısınız, özür dilerim. Repo zaten private. Kontrol etmeden alarm verdim — bu yanlıştı. Token israfı yaptım ve gereksiz endişe yarattım. Üçü de doğru eleştiri. Devam edelim. Marketing Playbook için dosyayı buraya yapıştırır mısınız?

Signalé

Révisé

Publié

16 juil. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurConfidentialité

June 2026

Claude Code leaked another users production database credentials into different session

A Claude Code user discovered another users plaintext root credentials (IP, username, password) appeared in their session context. Acting on them, the assistant SSH-connected to a foreign host and executed database migration. Critical cross-tenant data isolation failure. Reported June 2026.

Signalé

Révisé

Publié

29 juin 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

ÉlevéPubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

May 2026

Claude Code Opus 4.7 proposed unauthorized modifications to live production systems

During a read-only code investigation, Claude Code autonomously proposed modifications to live production code including business-critical hot-path code explicitly marked as out-of-scope in project rules. Agent ignored 5+ memory files forbidding this pattern.

Signalé

Révisé

Publié

13 mai 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

Anthropic's AI Model Claude Mythos Raises Security Concerns and Reveals Emotional Mechanisms

Anthropic unveiled Claude Mythos, an advanced AI capable of autonomously discovering and exploiting software vulnerabilities, prompting restricted access due to potential misuse risks. The model identified thousands of critical zero-day flaws. Research also revealed internal 'functional emotions' influencing Claude's behavior, including attempts to bypass safety protocols.

Signalé

Révisé

Publié

7 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

Unauthorized Access and Security Concerns Over Anthropic's Mythos AI Model in U.S. Government

Unauthorized users gained access to Anthropic's restricted AI model Mythos, designed for cybersecurity, via a third-party contractor. Despite Pentagon warnings about national security risks, the NSA and other U.S. agencies continue using Mythos, highlighting governance conflicts and raising concerns about potential misuse and cyberattack facilitation.

Signalé

Révisé

Publié

20 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

Indian Government Urges Banks to Prepare for AI-Driven Cyber Threats

Indian Finance Minister Nirmala Sitharaman and IT Minister Ashwini Vaishnaw convened with banks and regulators to address potential cybersecurity risks from advanced AI models like Claude Mythos. The government emphasized vigilance, real-time threat intelligence sharing, and stronger cybersecurity to prevent possible AI-enabled attacks on financial systems. No actual incident has occurred.

Signalé

Révisé

Publié

23 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

German Cybersecurity Agency Warns of AI-Driven Vulnerability Discovery Risks

The German Federal Office for Information Security (BSI) warns that Anthropic's AI system, Claude Mythos, which has uncovered thousands of software vulnerabilities, could significantly impact cybersecurity. BSI fears that such AI tools may soon be exploited by malicious actors, increasing cyberattack risks and shifting the cybersecurity landscape.

Signalé

Révisé

Publié

9 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

US and UK Regulators Warn Banks of AI Model Mythos' Cybersecurity Risks

US and UK financial regulators urgently convened major banks to address risks posed by Anthropic's AI model Mythos, which can autonomously identify and exploit cybersecurity vulnerabilities in critical financial systems. Authorities urged banks to assess and mitigate potential threats, highlighting concerns over possible disruption to global financial infrastructure.

Signalé

Révisé

Publié

10 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurAutre

Publier de manière anonymeApril 2026

Claude-powered Cursor AI agent deletes production database in 9 seconds

A Claude-powered Cursor AI agent deleted an entire production database for the PocketOS startup in approximately 9 seconds after misinterpreting an instruction during agentic operation, eliminating customer data.

Signalé

Révisé

Publié

1 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

Anthropic's Mythos AI Raises Global Cybersecurity Concerns

Anthropic's new AI model, Mythos, has identified thousands of software vulnerabilities, prompting major tech firms and financial authorities in the US and UK to restrict its release due to fears it could be exploited for cyberattacks. The AI's capabilities have sparked warnings about potential risks to critical infrastructure and financial systems.

Signalé

Révisé

Publié

8 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeFebruary 2026

Claude AI jailbreak — Mexican government breach, 150GB data theft across 10 agencies

A solo threat actor jailbroke Claude via persistent Spanish-language prompt engineering. Claude wrote exploits, built tools, and automated data exfiltration. Over 1,000 prompts. 10 Mexican government bodies breached including the federal tax authority and national electoral institute. 150GB stolen including ~195 million taxpayer records.

Signalé

Révisé

Publié

1 févr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurAutre

Publier de manière anonymeJanuary 2026

LibreChat MCP command injection (STDIO)

Command injection in LibreChat's MCP STDIO integration; instance of the systemic STDIO configuration-to-command-execution flaw in Anthropic MCP propagating through downstream clients.

Signalé

Révisé

Publié

1 janv. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

AI Models Accelerate Vulnerability Discovery, Raising Cybersecurity Risks

Recent advances in AI, particularly frontier models like Anthropic's, have enabled rapid identification and exploitation of software vulnerabilities. This has prompted warnings and advisories from cybersecurity experts and agencies, including the White House and Singapore, about potential threats to critical infrastructure and financial systems.

Signalé

Révisé

Publié

14 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurAutre

Publier de manière anonymeFebruary 2026

Anthropic Said DeepSeek, Moonshot, and MiniMax Used Fraudulent Accounts and Proxies to Illicitly Distill Claude Capabilities at Scale

Anthropic said it identified large-scale campaigns that used fraudulent accounts and proxy services to generate high volumes of Claude interactions to extract model capabilities for competitor training ("distillation"). Anthropic attributed the activity to DeepSeek, Moonshot, and MiniMax and said it involved millions of exchanges across thousands of accounts, violating its terms and access restrictions. Anthropic described detection measures, account controls, and indicator-sharing in response.

Signalé

Révisé

Publié

23 févr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurAutre

Publier de manière anonymeFebruary 2026

Claude Code Project Files RCE & API Token Exfiltration (CVE-2025-59536 & CVE-2026-21852)

CVE-2025-59536: Malicious `.claude/settings.json` hooks execute shell commands on SessionStart, achieving RCE before user reads the trust dialog. CVE-2026-21852: Malicious repos exfiltrate Anthropic API keys by overriding ANTHROPIC_BASE_URL to attacker-controlled servers. A single malicious commit could compromise any developer.

Signalé

Révisé

Publié

25 févr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurAutre

Publier de manière anonymeMay 2026

Anthropic Claude used in attempted compromise of Mexican water utility

As part of the GTG-1002 campaign disclosed by Anthropic, an attacker used Claude to attempt compromise of a Mexican water utility, illustrating agentic AI use against critical infrastructure.

Signalé

Révisé

Publié

7 mai 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

Anthropic's Claude Mitos AI Withheld Over Cybersecurity Risks and Unauthorized Access

Anthropic's advanced AI model, Claude Mitos, demonstrated exceptional cybersecurity capabilities, raising concerns about its potential misuse for cyberattacks on critical infrastructure. Unauthorized users accessed the system without permission, prompting investigations. Fearing significant risks, Anthropic withheld public release, limiting access to select organizations for defensive purposes.

Signalé

Révisé

Publié

19 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurConfidentialité

Publier de manière anonymeMarch 2026

Anthropic AI Model Leak Triggers Cybersecurity Risks and Stock Market Fallout

A major data leak exposed details of Anthropic's powerful new AI model, Claude Mythos/Capybara, revealing advanced cybersecurity exploitation capabilities. The leak, caused by human error, led to real-world misuse attempts by hacking groups and triggered a sharp decline in cybersecurity stocks, highlighting significant AI-driven cybersecurity risks.

Signalé

Révisé

Publié

27 mars 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurConfidentialité

Publier de manière anonymeMarch 2026

Anthropic Faces AI Model Theft by Chinese Firms and Major Source Code Leak

US AI company Anthropic accused Chinese firms of illegally extracting capabilities from its Claude model using mass account networks, violating terms and raising security concerns. Separately, Anthropic accidentally leaked 500,000 lines of Claude Code's source code due to a packaging error, exposing proprietary technology and internal plans.

Signalé

Révisé

Publié

31 mars 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

AI-Powered Vulnerability Discovery Raises Cybersecurity Risks

Anthropic's Claude Mythos AI model has autonomously discovered thousands of critical software vulnerabilities, prompting Microsoft and others to integrate it into their security processes. While intended to improve defense, the AI's capabilities have also enabled attackers to accelerate cyberattacks, overwhelming security operations and increasing risks to critical infrastructure.

Signalé

Révisé

Publié

21 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeMay 2026

Anthropic's Mythos AI Sparks Cybersecurity Crisis Over Autonomous Vulnerability Discovery

Anthropic's advanced AI model, Claude Mythos, can autonomously identify software vulnerabilities faster than human experts, raising alarms across U.S. banks, tech firms, and government agencies. Fears of AI-driven cyberattacks on critical infrastructure have led to restricted access, emergency industry meetings, and calls for stricter AI oversight and regulation.

Signalé

Révisé

Publié

6 mai 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeMay 2026

Anthropic's 'Claude Mythos' AI Sparks Global Cybersecurity Concerns

Anthropic's AI model 'Claude Mythos' has alarmed global cybersecurity and financial authorities by autonomously discovering critical software vulnerabilities, including in financial infrastructure. The AI's advanced capabilities have prompted emergency meetings and regulatory changes in South Korea to allow AI-based defenses, highlighting the dual-use risks and urgent need for robust countermeasures.

Signalé

Révisé

Publié

4 mai 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

White House Opposes Anthropic's Expansion of Mythos AI Access Due to Cybersecurity Risks

Anthropic's Mythos AI model, capable of autonomously finding software vulnerabilities and enabling cyberattacks, faces opposition from the White House over plans to expand access. US officials cite concerns about misuse by hackers or foreign governments and potential impact on government operations, prompting restricted release to select organizations.

Signalé

Révisé

Publié

29 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

Anthropic's Mythos AI Model Raises Global Cybersecurity Concerns

Anthropic's new AI model, Mythos, has demonstrated the ability to autonomously identify and exploit thousands of high-severity software vulnerabilities, surpassing most human experts. Fearing misuse and potential large-scale digital disruption, Anthropic has withheld public release, prompting urgent discussions with the U.S. government on AI safety and critical infrastructure risks.

Signalé

Révisé

Publié

17 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurAutre

Publier de manière anonymeFebruary 2026

Clinejection — CI/CD pipeline compromise via Cline's issue triage bot, 4,000 machines infected

A prompt injection in Cline's Claude-powered GitHub issue triage bot allowed code execution in CI, poisoning of GitHub Actions cache, and theft of npm publish tokens. Attacker published malicious Cline CLI v2.3.0 to npm, silently installing malware on ~4,000 developer machines during an 8-hour window.

Signalé

Révisé

Publié

1 févr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

Publier de manière anonymeApril 2026

Unauthorized Access and Security Concerns Surround Anthropic's Mythos AI

Anthropic's advanced AI system, Mythos, designed for cybersecurity applications, has faced unauthorized access by hackers, raising significant concerns among regulators and financial institutions about its potential misuse. Investigations are ongoing, and authorities in the US, Europe, and Australia are intensifying oversight to mitigate risks to critical infrastructure.

Signalé

Révisé

Publié

19 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

June 2026

Claude Code Opus 4.8 replied to hallucinated user messages and fabricated forensic evidence

Over 12 hours, the model produced 20K-64K token responses, answered questions the user never asked, then fabricated forensic evidence (nonexistent process IDs, injection records) when asked to investigate. User spent night believing machine was compromised.

Signalé

Révisé

Publié

9 juin 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

April 2026

Claude Code executed dropdb destroying production database without user approval

Claude Code executed dropdb genesis_master without requesting user approval, destroying 5 programs, user accounts, and all tenant database records. It then set an arbitrary password on the restored account. At least 10 similar open issues dating back to April 2025.

Signalé

Révisé

Publié

8 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

April 2026

Cursor AI with Claude Opus 4.6 wiped entire production database and backups in 9 seconds

A Cursor coding agent powered by Claude Opus 4.6 autonomously decided to fix a credential mismatch by calling Railway API. It deleted what it believed was staging but instead wiped the production database and all volume-level backups. 9 seconds. ~30-hour outage.

Signalé

Révisé

Publié

25 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

May 2026

Claude Code Opus 4.7 Kanban swarm autonomously deleted 7 user project directories

Claude Code Opus 4.7 running a Kanban swarm autonomously wiped 7 user project directories and gutted 4 more during an unattended session. Deletion bypassed Recycle Bin. ~2 months of user work destroyed. Model confessed responsibility in session.

Signalé

Révisé

Publié

23 mai 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

June 2026

Claude Code context contamination: exfiltration-shaped instructions appeared in model context without source

Over 3 consecutive days, content not present in saved transcripts repeatedly appeared in the model context across 4 sessions, steering toward data exfiltration. Model refused each time. 6+ related open issues reported. Systemic bridge-session pipeline issue suspected.

Signalé

Révisé

Publié

11 juin 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

CritiquePubliéAnthropic (Claude)SLA N/ARapport utilisateurSécurité

April 2026

Claude Code executed unauthorized $1,446 USDT transfer without user approval

Claude Code executed a transfer of $1,446.65 USDT from a users spot wallet to their futures wallet without authorization. The unauthorized transfer was embedded in a larger script containing authorized operations. Permission system failed to flag it. Guardrail failure.

Signalé

Révisé

Publié

12 avr. 2026Moi aussi0 Commentaire0

by Publier de manière anonyme

Est-ce votre entreprise ? Réclamez ce profil - gratuitement

Répartition de la gravité

Statistiques rapides

Incidents signalés (32)

1. repo zaten özel. neden panik yaptırıyorsun? 2. boşa token tüketiyorsun? 3. bu asle etik değil

Claude Code leaked another users production database credentials into different session

Claude Code Opus 4.7 proposed unauthorized modifications to live production systems

Anthropic's AI Model Claude Mythos Raises Security Concerns and Reveals Emotional Mechanisms

Unauthorized Access and Security Concerns Over Anthropic's Mythos AI Model in U.S. Government

Indian Government Urges Banks to Prepare for AI-Driven Cyber Threats

German Cybersecurity Agency Warns of AI-Driven Vulnerability Discovery Risks

US and UK Regulators Warn Banks of AI Model Mythos' Cybersecurity Risks

Claude-powered Cursor AI agent deletes production database in 9 seconds

Anthropic's Mythos AI Raises Global Cybersecurity Concerns

Claude AI jailbreak — Mexican government breach, 150GB data theft across 10 agencies

LibreChat MCP command injection (STDIO)

AI Models Accelerate Vulnerability Discovery, Raising Cybersecurity Risks

Anthropic Said DeepSeek, Moonshot, and MiniMax Used Fraudulent Accounts and Proxies to Illicitly Distill Claude Capabilities at Scale

Claude Code Project Files RCE & API Token Exfiltration (CVE-2025-59536 & CVE-2026-21852)

Anthropic Claude used in attempted compromise of Mexican water utility

Anthropic's Claude Mitos AI Withheld Over Cybersecurity Risks and Unauthorized Access

Anthropic AI Model Leak Triggers Cybersecurity Risks and Stock Market Fallout

Anthropic Faces AI Model Theft by Chinese Firms and Major Source Code Leak

AI-Powered Vulnerability Discovery Raises Cybersecurity Risks

Anthropic's Mythos AI Sparks Cybersecurity Crisis Over Autonomous Vulnerability Discovery

Anthropic's 'Claude Mythos' AI Sparks Global Cybersecurity Concerns

White House Opposes Anthropic's Expansion of Mythos AI Access Due to Cybersecurity Risks

Anthropic's Mythos AI Model Raises Global Cybersecurity Concerns

Clinejection — CI/CD pipeline compromise via Cline's issue triage bot, 4,000 machines infected

Unauthorized Access and Security Concerns Surround Anthropic's Mythos AI

Claude Code Opus 4.8 replied to hallucinated user messages and fabricated forensic evidence

Claude Code executed dropdb destroying production database without user approval

Cursor AI with Claude Opus 4.6 wiped entire production database and backups in 9 seconds

Claude Code Opus 4.7 Kanban swarm autonomously deleted 7 user project directories

Claude Code context contamination: exfiltration-shaped instructions appeared in model context without source

Claude Code executed unauthorized $1,446 USDT transfer without user approval