Loading story…
Cybersecurity threats leveraging artificial intelligence are rapidly evolving, with attackers using AI to automate vulnerability discovery, conduct deepfake-enabled social engineering, and compromise large language models (LLMs) through data poisoning and malicious model files. A Gartner survey found that 62% of organizations experienced deepfake attacks, often combined with social engineering to defraud companies, while 32% faced attacks on AI applications such as prompt injection. Security researchers have demonstrated prompt injection vulnerabilities in AI summarization tools, highlighting the risks of insufficient separation between user input and system prompts. Additionally, experts warn that LLMs can be compromised via poisoned training data, malicious adapters, or trojanized model files, expanding the attack surface for organizations adopting AI technologies. These developments underscore the need for robust AI governance, technical defenses, and employee awareness to counter increasingly sophisticated AI-driven threats.
Mallory correlates global threat intelligence with your attack surface — know if you’re exposed before adversaries strike.
4 events from the most recent confirmed update back to the earliest known activity.
An Infosec Writeups article described how attackers can use AI to find vulnerabilities faster. The publication reflects public discussion of AI-enabled offensive security techniques.
An Infosec Writeups post detailed the discovery of a prompt injection issue in an AI summarization workflow, where a hidden note could mislead the summarizer. The article constitutes a public disclosure of the bug bounty finding and its attack technique.
SC Media reported Gartner findings that 62% of firms had been hit by deepfake attacks. This represents a new data point on the scale of deepfake-related security incidents affecting organizations.
Trend Micro released a research article titled "This Is How Your LLM Gets Compromised," outlining ways large language models can be attacked or manipulated. The publication marks a technical disclosure of LLM security risks and compromise methods.
4 references tracked. Mallory keeps watching after this page renders.
infosecwriteups.com
Open sourceinfosecwriteups.com
Open sourcescworld.com
Open sourcetrendmicro.com
Open sourceMap indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.
Other stories Mallory is tracking that overlap on entities, sources, or topic labels.
A recent Gartner survey revealed that 62% of organizations have experienced deepfake attacks within the past year, highlighting the rapid adoption of AI-driven social engineering tactics. These attacks often involve the use of deepfake technology to impersonate executives, tricking employees into transferring funds or divulging sensitive information. Akif Khan of Gartner emphasized that social engineering remains a reliable attack vector, and the introduction of deepfakes makes it even more challenging for employees to detect fraudulent activity. Automated defenses alone are insufficient, as employees are now the frontline defense against these sophisticated impersonation attempts. The survey also found that 32% of organizations faced attacks targeting AI applications, particularly through prompt injection and manipulation of large language models (LLMs). Such adversarial prompting can cause AI chatbots and assistants to generate biased or malicious outputs, further expanding the threat landscape. Flashpoint analysts corroborate these findings, reporting that threat actors are actively discussing and deploying AI-powered tools in underground communities. These include specialized malicious AI models and AI-generated attack plans, which are being used to automate and scale cybercriminal operations. The most immediate threat identified is the use of AI to exploit human psychology, with attackers leveraging AI to create convincing phishing lures and fabricated realities that undermine traditional authentication methods based on voice and visual cues. Financial institutions are particularly vulnerable, as demonstrated by recent incidents where finance workers were deceived by AI-generated content. The rise of 'Dark GPTs' and Attack-as-a-Service (AaaS) offerings on the dark web further illustrates the commercialization and accessibility of AI-driven cybercrime. Security experts recommend a defense-in-depth approach, combining robust technical controls with targeted measures for emerging AI risks. AI-powered security awareness training is increasingly seen as essential, empowering employees to recognize and resist sophisticated social engineering attacks. Over 70,000 organizations are already leveraging such platforms to strengthen their human firewall. As generative AI adoption accelerates, organizations must remain vigilant against both direct deepfake attacks and indirect threats to AI application infrastructure. The evolving threat landscape demands continuous adaptation of security strategies to address the growing use of AI in cybercrime. Proactive threat intelligence and employee education are critical components in mitigating these risks. Organizations are urged to avoid isolated investments and instead implement comprehensive controls tailored to each new category of AI-driven threat. The convergence of deepfake technology, AI-powered phishing, and prompt-based attacks marks a significant escalation in the sophistication and scale of cyber threats facing enterprises today.
Jun 29, 2026
The rapid adoption of AI agents and large language models (LLMs) by software developers is transforming the software development pipeline, increasing productivity but also introducing significant security risks. As organizations integrate AI tools for code generation, debugging, and architectural design, the quality and security of code have become inconsistent, with vulnerabilities in legacy code often being propagated. Experts warn that while AI can enhance bug detection and triage, the sheer volume and complexity of AI-generated code may outpace human oversight, making it easier for insecure code to reach production. Additionally, the use of AI in privileged access management is expected to shift from passive monitoring to proactive, autonomous governance, with machine learning models enforcing real-time policies and detecting anomalous behavior to prevent insider threats and account takeovers. The evolving threat landscape is further complicated by attackers leveraging AI-powered tools and deepfakes to conduct sophisticated scams and social engineering campaigns. For example, the Nomani investment scam has surged by 62%, using AI-generated video testimonials and deepfake ads on social media to deceive victims. Security researchers also highlight the abuse of legitimate open-source tools and the use of synthetic data in cyber deception, as well as the need for organizations to address the growing trust gap in AI technologies. As AI becomes more deeply embedded in both offensive and defensive cybersecurity operations, organizations must prioritize secure development practices, adaptive authentication, and continuous monitoring to mitigate emerging risks.
Jun 29, 2026
Cybersecurity professionals are rapidly adopting artificial intelligence (AI) tools to enhance threat detection, investigation, and response, with over 90% of surveyed teams now testing or planning to use AI in their operations. However, this widespread adoption brings new security challenges, as highlighted by recent research and industry reports. The Cloud Security Alliance and Google Cloud emphasize that traditional data security models require significant updates to address AI-specific risks such as prompt injection, model inversion, and multi-modal data leakage. Unlike conventional vulnerabilities, prompt injection exploits the inherent ambiguity of large language models (LLMs), making it a persistent risk that cannot be mitigated by simple patches. Security experts recommend combining AI-driven analysis with deterministic, auditable controls to ensure reliable and explainable security decisions, especially in enforcement actions like access revocation or incident response. A concrete example of these risks was demonstrated in Docker's 'Ask Gordon' AI assistant, where researchers exploited a metadata-based prompt injection flaw to exfiltrate sensitive information. Attackers could embed malicious instructions in the metadata of Docker Hub repositories, which the AI would then execute when prompted by users, highlighting the real-world impact of prompt injection vulnerabilities. The evolving threat landscape also includes the use of malicious LLMs and AI-powered tools in DDoS-for-hire operations, with underground actors leveraging AI to automate botnet recruitment and evade detection. These developments underscore the urgent need for organizations to update their security frameworks, implement ongoing risk management for AI systems, and remain vigilant against emerging AI-driven attack vectors.
Jun 29, 2026