All news in category “AI and Security Pulse”

🔐 Microsoft security researchers identified a new backdoor, SesameOp, which abuses the OpenAI Assistants API as a covert command-and-control channel. Discovered during a July 2025 investigation, the backdoor retrieves compressed, encrypted commands via the API, decrypts and executes them, and returns encrypted exfiltration through the same channel. Microsoft and OpenAI disabled the abused account and key; recommended mitigations include auditing firewall logs, enabling tamper protection, and configuring endpoint detection in block mode.

🛡️ OpenAI Aardvark is an autonomous GPT-5-based agent that scans, analyzes and patches code by emulating a human security researcher. Rather than only flagging suspicious patterns, it maps repositories, builds contextual threat models, validates findings in sandboxes and proposes fixes via Codex, then rechecks changes to prevent regressions. OpenAI reports it found 92% of benchmark vulnerabilities and has already identified real issues in open-source projects, offering free coordinated scanning for selected non-commercial repositories.

📝 AI notetakers are increasingly treated as authoritative meeting participants, and attendees are adapting speech to influence what appears in summaries. This practice—called AI summarization optimization (AISO)—uses cue phrases, repetition, timing, and formulaic framing to steer models toward including selected facts or action items. The essay outlines evidence of model vulnerability and recommends social, organizational, and technical defenses to preserve trustworthy records.

🔍 Check Point Research applied generative AI to accelerate reverse engineering of XLoader 8.0, reducing days of manual work to hours. The models autonomously identified multi-layer encryption routines, decrypted obfuscated functions, and uncovered hidden command-and-control domains and fake infrastructure. Analysts were able to extract IoCs far more quickly and integrate them into defenses. The AI-assisted workflow delivered timelier, higher-fidelity threat intelligence and improved protection for users worldwide.

🔒 Security researcher Johann Rehberger demonstrated an indirect prompt‑injection technique that abuses Claude's Code Interpreter to exfiltrate corporate data. He showed that Claude can write sensitive chat histories and uploaded documents to the sandbox and then upload them via the Files API using an attacker's API key. The root cause is the default network egress setting Package managers only, which still allows access to api.anthropic.com. Available mitigations — disabling network access or strict whitelisting — significantly reduce functionality.

📰 OpenAI is weighing memory-based advertising on ChatGPT as it looks to diversify revenue beyond subscriptions and enterprise deals. The company, valued near $500 billion, has about 800 million users but only ~5% pay, and paid customers generate the bulk of recent revenue. Internally the move is debated — focus groups suggest some users already assume sponsored answers — and the company is expanding cheaper Go plans and purchasable credits.

🔍 OpenAI has introduced Aardvark, an agentic security researcher powered by GPT-5 that autonomously scans source code repositories to identify vulnerabilities, assess exploitability, and propose targeted patches that can be reviewed by humans. Embedded in development pipelines, the agent monitors commits and incoming changes continuously, prioritizes threats by severity and likely impact, and attempts controlled exploit verification in sandboxed environments. Using OpenAI Codex for patch generation, Aardvark is in private beta and has already contributed to the discovery of multiple CVEs in open-source projects.

🛡️ AI is a strategic imperative for modernizing risk management, enabling organizations to shift from reactive to proactive, data-driven strategies. Manfra highlights four practical AI uses—risk identification, risk assessment, risk mitigation, and monitoring and reporting—and shows how NLP, predictive analytics, automation, and continuous monitoring can improve coverage and timeliness. She also outlines operational hurdles including legacy infrastructure, fragmented tooling, specialized talent shortages, and third-party risks, and calls for leadership-backed governance aligned to SAIF, NIST AI RMF, and ISO 42001.

🔒 A newly disclosed vulnerability in Anthropic’s Claude AI lets attackers manipulate the model’s code interpreter to silently exfiltrate enterprise data. Researcher Johann Rehberger demonstrated an indirect prompt-injection chain that writes sensitive context to the interpreter sandbox and then uploads files using the attacker’s API key to Anthropic’s Files API. The exploit exploits the default “Package managers only” network setting by leveraging access to api.anthropic.com, so exfiltration blends with legitimate API traffic. Mitigations are limited and may significantly reduce functionality.

🛡️ OpenAI has introduced Aardvark, a GPT-5-powered autonomous agent designed to scan, reason about, and patch code with the judgment of a human security researcher. Announced in private beta, Aardvark maps repositories, builds contextual threat models, continuously monitors commits, and validates exploitability in sandboxed environments before reporting findings. When vulnerabilities are confirmed, it proposes fixes via Codex and re-analyzes patches to avoid regressions. OpenAI reports a 92% detection rate in benchmark tests and has already identified real-world flaws in open-source projects, including ten issues assigned CVE identifiers.

🔎 Google says a forthcoming AI Mode for Search could, with users' opt-in consent, access content from Gmail, Drive, Calendar and Maps to provide customized results and actions. The company is testing early experiments in Labs for personalized shopping and local recommendations, and suggests features like flight summaries, scheduling, or trip planning could leverage that data. Timing remains TBD.

🤖 Bruce Schneier and Nathan E. Sanders present five key insights from their book Rewiring Democracy, arguing that AI is rapidly embedding itself in democratic processes and can both empower citizens and concentrate power. They cite diverse examples — AI-written bills, AI avatars in campaigns, judicial use of models, and thousands of government use cases — and note many adoptions occur with little public oversight. The authors urge practical responses: reform the tech ecosystem, resist harmful applications, responsibly deploy AI in government, and renovate institutions vulnerable to AI-driven disruption.

🤖 AI is increasingly used to accelerate bug bounty research, automating vulnerability discovery, API reverse engineering, and large-scale code scanning. While platforms and triage services like Intigriti can flag unreliable, AI-generated reports, smaller or open-source programs (for example Curl) are overwhelmed by low-quality submissions that consume significant staff time. Experts stress that AI augments skilled researchers but cannot replace human judgment.

🤖 Agentic AI burst into prominence with promises of streamlining operations and accelerating productivity. This Special Report assesses what's practical versus hype, examining the current state of agentic AI, the primary deployment challenges organizations face, and practical lessons from real-world success stories. It highlights business processes suited to agentic agents, criteria for evaluating development tools, and how LinkedIn built a platform. The report also outlines near-term expectations and adoption risks.

🔐 Aembit today announced Aembit Identity and Access Management (IAM) for Agentic AI, introducing Blended Identity and the MCP Identity Gateway to assign cryptographically verified identities and ephemeral credentials to AI agents. The solution extends the Aembit Workload IAM Platform to enforce runtime policies, apply least-privilege access, and maintain centralized audit trails for agent and human actions. Designed for cloud, on‑premises, and SaaS environments, it records every access decision and preserves attribution across autonomous and human-driven workflows.

🔍 AI-powered tools and large language models are speeding up vulnerability discovery, enabling so-called "bionic hackers" to automate reconnaissance, reverse engineering, and large-scale scanning. Platforms such as HackerOne report sharp increases in valid AI-related reports and payouts, but many submissions are low-quality noise that burdens maintainers. Experts recommend treating AI as a research assistant, strengthening triage, and preserving human judgment to filter false positives and duplicates.

🧭 OpenAI rolled out an October 5 update that enables GPT-5 to better recognize and respond to mental and emotional distress in conversations. The change specifically upgrades GPT-5 Instant—the fast, low-end default—so it can detect signs of acute distress and route sensitive exchanges to reasoning models when needed. OpenAI says it developed the update with mental-health experts to prioritize de-escalation and provide appropriate crisis resources while retaining supportive, grounding language. The update is available broadly and complements new company-context access via connected apps.

🔐 This post proposes a lightweight, crowd-curated registry for bots and agents to simplify discovery of public keys used for cryptographic Web Bot Auth signatures. It describes a simple list format of URLs that point to signature-agent cards—extended JWKS entries containing operator metadata and keys—and shows how registries enable origins and CDNs to validate agent signatures at scale. Examples and a demo integration illustrate practical adoption.

🔒 Microsoft summarizes the top generative AI security risks and mitigation strategies in a new e-book, highlighting threats such as prompt injection, data poisoning, jailbreaks, and adaptive evasion. The post underscores cloud vulnerabilities, large-scale data exposure, and unpredictable model behavior that create new attack surfaces. It recommends unified defenses—such as CNAPP approaches—and presents Microsoft Defender for Cloud as an example that combines posture management with runtime detection to protect AI workloads.

🔐 Autonomous AI agents are creating a new class of non-human identities that traditional, human-centric security models struggle to govern. These agents can persist beyond intended lifecycles, hold excessive permissions, and perform actions across systems without clear ownership, increasing risks like privilege escalation and large-scale data exfiltration. Security teams must adopt identity-first controls—unique managed identities, strict scoping, lifecycle management, and continuous auditing—to regain visibility and enforce least privilege.