# Social engineering Published articles for Social engineering. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Revolut phishing texts appear days after data breach DevFeed: [Revolut phishing texts appear days after data breach]() Original publisher: [Read original article]() Author: Pieter Arntz Published: 2026-09-17T14:07:15Z Content type: news Language: en Sources: [Malwarebytes]() Topics: [Social engineering](), [passwords](), [VirusTotal]() Tags: [account](), [breach](), [password](), [phishing](), [revolut](), [scam](), [scams](), [social-engineering](), [threat-intel]() ### AI overview Revolut customers received phishing texts days after the bank disclosed customer data to a government impostor. The report says the campaign's connection to the breach is not yet known and describes a fake identity check designed to obtain passwords. ### Source excerpt Revolut customers received phishing texts only days after the digital bank acknowledged disclosing customer data to a government impostor. ## HBO Max's verified Reddit account hijacked to spread malware DevFeed: [HBO Max's verified Reddit account hijacked to spread malware]() Original publisher: [Read original article]() Author: Pieter Arntz Published: 2026-09-15T11:51:03Z Content type: news Language: en Sources: [Malwarebytes]() Topics: [Malware](), [ClickFix](), [Reddit](), [Social engineering](), [passwords](), [Cryptocurrency](), [macOS](), [Windows]() Tags: [clickfix](), [cryptocurrency](), [hbo-max](), [macos](), [malware](), [news](), [passwords](), [pasteswitch](), [powershell](), [reddit](), [security](), [social-engineering](), [terminal](), [windows]() ### AI overview Cybercriminals hijacked HBO Max's verified Reddit account and used it to run 108 malicious ads over about 48 hours. The ads promoted fake software and used ClickFix-style instructions to distribute infostealers and cryptocurrency clipboard hijackers to macOS and Windows users. ### Source excerpt Cybercriminals used HBO Max's verified Reddit account to run 108 malicious ads that tricked people into installing information stealers. ## Protecting organizations from AI-assisted executive impersonation and invoice fraud DevFeed: [Protecting organizations from AI-assisted executive impersonation and invoice fraud]() Original publisher: [Read original article]() Author: Microsoft Security Research Published: 2026-09-10T17:23:05Z Content type: article Language: en Sources: [Microsoft Security Blog]() Topics: [spoofing]() Tags: [ai](), [enterprise](), [executive](), [fraud](), [generative-ai](), [scam](), [security](), [social-engineering]() ### AI overview Microsoft examines an email-fraud campaign that used AI-assisted executive impersonation, fabricated invoices, and fake supporting conversations to prompt ACH payments. ### Source excerpt Microsoft examines an AI-assisted business email compromise campaign that used executive impersonation and fake invoices to target finance teams with ACH payment fraud. The post Protecting organizations from AI-assisted executive impersonation and invoice fraud appeared first on Microsoft Security Blog. ## Detect and disrupt AI-themed attacks with Microsoft Defender DevFeed: [Detect and disrupt AI-themed attacks with Microsoft Defender]() Original publisher: [Read original article]() Author: Rob Lefferts Published: 2026-09-10T16:00:00Z Content type: article Language: en Sources: [Microsoft Security Blog]() Topics: [Malware](), [Threat Research](), [spoofing](), [Microsoft]() Tags: [adversary-in-the-middle-aitm](), [ai](), [attacks](), [chatgpt](), [claude](), [copilot](), [credential-theft](), [deepseek](), [defender](), [github](), [malware](), [microsoft](), [phishing](), [security](), [social-engineering]() ### AI overview Microsoft describes AI-themed phishing, malvertising, credential theft, and malware campaigns that impersonate popular AI services and tools. It argues that attackers are exploiting trust and urgency around AI brands rather than compromising the referenced services. ### Source excerpt See how Microsoft Defender detects and disrupts AI-themed phishing, malware, and multi-stage attacks across the attack chain. The post Detect and disrupt AI-themed attacks with Microsoft Defender appeared first on Microsoft Security Blog. ## Passkey-themed social engineering leads to identity and cloud compromise DevFeed: [Passkey-themed social engineering leads to identity and cloud compromise]() Original publisher: [Read original article]() Author: Microsoft Security Research, Krithika Ramakrishnan, Bharat Vaghela, Vaibhav Deshmukh, Subhajit Ghosh, Anusha Chakraborty, Akash Chaudhuri, Victor Chingtham and Ivan Macalintal Published: 2026-09-09T17:41:18Z Content type: article Language: en Sources: [Microsoft Security Blog]() Topics: [MFA](), [data-processing]() Tags: [adversary-in-the-middle-aitm](), [apis](), [attacks](), [authentication](), [cloud](), [data](), [identity](), [mfa](), [phishing](), [security](), [social-engineering]() ### AI overview Microsoft Security Research describes a passkey-themed social-engineering campaign that compromises cloud identities through AiTM phishing or device-code flows, establishes authentication persistence, and collects cloud data. It outlines investigation signals and recommends revoking sessions and removing unauthorized authentication methods after confirmed compromise. ### Source excerpt Passkey-themed social engineering is being used to compromise identities and enable broader cloud attacks. Learn how threat actors establish MFA persistence, abuse Microsoft Graph for reconnaissance, and access SharePoint, OneDrive, and email data, along with key detection and mitigation guidance. The post Passkey-themed social engineering leads to identity and cloud compromise appeared first on Microsoft Security Blog. ## X Money rollout linked to password-reset attacks DevFeed: [X Money rollout linked to password-reset attacks]() Original publisher: [Read original article]() Author: Pieter Arntz Published: 2026-09-04T12:29:41Z Content type: news Language: en Sources: [Malwarebytes]() Topics: [passwords]() Tags: [account-takeover](), [attacks](), [breach](), [financial-services](), [how-to](), [news](), [password-reset](), [scams](), [security](), [social-engineering](), [x-money]() ### AI overview X is investigating unsolicited password-reset emails sent to users amid the wider availability of X Money. The company says it has found no evidence of a breach, successful account takeovers, or access to X Money funds. ### Source excerpt As X expands into payments, users are receiving password-reset emails they didn't request. Here's what may be happening and how to stay safe. ## ASCII smuggling crosses over from AI prompt injection to phishing evasion DevFeed: [ASCII smuggling crosses over from AI prompt injection to phishing evasion]() Original publisher: [Read original article]() Author: Microsoft Security Research, Noam Kochavi and Sarah Wolstencroft Published: 2026-09-03T16:00:00Z Content type: article Language: en Sources: [Microsoft Security Blog]() Topics: [ASCII](), [Machine Learning, Security Attacks](), [Language models](), [telemetry](), [Microsoft]() Tags: [ai-models](), [ai-security](), [ascii](), [microsoft](), [phishing](), [security](), [social-engineering](), [telemetry]() ### AI overview Microsoft describes a phishing campaign that uses invisible Unicode tag characters to split lure words and evade email parsing. The technique, known as ASCII smuggling, was previously prominent in AI prompt-injection research because models can process hidden text that people cannot see. ### Source excerpt Invisible Unicode characters popularized for hiding instructions from AI models are now being used to obfuscate words before email filters parse them. The post ASCII smuggling crosses over from AI prompt injection to phishing evasion appeared first on Microsoft Security Blog. ## Impersonating IT support: how threat actors turn a remote session into enterprise-wide access DevFeed: [Impersonating IT support: how threat actors turn a remote session into enterprise-wide access]() Original publisher: [Read original article]() Author: Microsoft Security Research, Sagar Patil, Arlette Umuhire Sangwa, Jesse Birch and Ravikant Tiwari Published: 2026-09-02T22:51:18Z Content type: article Language: en Sources: [Microsoft Security Blog]() Topics: [High Profile Threats](), [Node.js]() Tags: [c2](), [identity](), [infrastructure](), [javascript](), [microsoft-teams](), [node-js](), [powershell](), [security](), [social-engineering](), [windows]() ### AI overview Microsoft analyzes an intrusion campaign in which attackers impersonate IT support through Microsoft Teams, obtain remote access, deploy a Node.js and JavaScript implant, and move laterally through enterprise systems. The article provides detection, mitigation, and hunting guidance. ### Source excerpt Microsoft Threat Intelligence observed a human-operated intrusion campaign that abuses Microsoft Teams external collaboration to impersonate IT support, gain remote access, and deploy a Node.js-based implant. Learn how attackers move from social engineering to lateral movement using legitimate tools, and how Microsoft Defender helps detect and disrupt the activity. The post Impersonating IT support: how threat actors turn a remote session into enterprise-wide access appeared first on Microsoft Security Blog. ## MAST in the Age of Open Source Software |Guardsquare DevFeed: [MAST in the Age of Open Source Software |Guardsquare]() Original publisher: [Read original article]() Author: Simon Haven - Product Marketing Manager Published: 2026-09-01T10:59:01Z Content type: article Language: en Sources: [Guardsquare Blog]() Topics: [Open Source](), [Mobile Security](), [Application Security](), [Vulnerabilities](), [Social engineering](), [npm]() Tags: [app-security-testing](), [application-security](), [appsweep](), [general](), [malware](), [mobile](), [mobile-apps](), [npm](), [open-source](), [open-source-software](), [social-engineering](), [supply-chain](), [vulnerabilities]() ### AI overview This article explains why mobile application security testing is important when apps use free and open source software. It describes how vulnerable, deprecated, compromised, and transitive third-party dependencies can expand an app's attack surface, citing a September 2025 npm supply chain attack as an example. ### Source excerpt Building with free and open source software (FOSS) has become common practice for app developers. In 2022, it was estimated that between 70% and 90% of any given software codebase was made up of open source components. Leveraging open source projects presents many advantages: ## Defending against AI-fueled social engineering DevFeed: [Defending against AI-fueled social engineering]() Original publisher: [Read original article]() Author: Joe DeFever Published: 2026-09-01T00:00:00Z Content type: article Language: en Sources: [Elastic Blog - Elasticsearch, Kibana, and ELK Stack]() Topics: [Social engineering](), [spoofing]() Tags: [agentic-ai-alerting-security-analytics](), [ai](), [attacks](), [phishing](), [security](), [soc](), [social-engineering]() ### AI overview The article describes how AI enables personalized spear phishing, smishing, and deepfake-based impersonation at greater scale. It argues that defenders should detect behavioral signals rather than rely solely on static signatures and known patterns. ### Source excerpt AI is supercharging social engineering. Learn how SOC teams can detect deepfakes, spear phishing, and smishing before they cause real damage. ## Spring Ring: An Inside Look at Voice Phishing Campaigns in Microsoft Teams DevFeed: [Spring Ring: An Inside Look at Voice Phishing Campaigns in Microsoft Teams]() Original publisher: [Read original article]() Author: Noam Sala Published: 2026-08-31T10:00:36Z Content type: article Language: en Sources: [Unit 42]() Topics: [Social engineering](), [Malware](), [remote access software](), [Microsoft](), [Cloaked Ursa](), [Entra ID](), [Incident response]() Tags: [attacks](), [cloaked-ursa](), [enterprise](), [entra-id](), [incident-response](), [malware](), [microsoft](), [microsoft-teams](), [payload](), [phishing](), [powershell](), [remote-access-trojan](), [social-engineering](), [spoof](), [threat-research](), [vishing](), [voice]() ### AI overview Spring Ring is a coordinated social engineering campaign that used external Microsoft Teams accounts and voice phishing to impersonate IT help desk staff. The operation targeted more than 150 employees across at least 10 companies and attempted to deliver remote monitoring and management tools or custom malware. A more advanced variant escalated to an NTLM relay attack against an organization's domain controller. ### Source excerpt Learn how the Spring Ring campaign abuses Microsoft Teams and voice phishing to deploy malware and target enterprise domain controllers. The post Spring Ring: An Inside Look at Voice Phishing Campaigns in Microsoft Teams appeared first on Unit 42. ## AI-driven OSINT in the wrong hands - and why everyone could be a target for fraud DevFeed: [AI-driven OSINT in the wrong hands - and why everyone could be a target for fraud]() Original publisher: [Read original article]() Author: Phil Muncaster Published: 2026-08-27T09:00:00Z Content type: article Language: en Sources: [WeLiveSecurity]() Topics: [Artificial Intelligence](), [Reconnaissance](), [Social engineering](), [Web](), [Malware]() Tags: [ai](), [attacks](), [fraud](), [malware](), [phishing](), [privacy](), [research](), [social-engineering](), [vulnerability]() ### AI overview AI-powered OSINT is making it faster and easier for cybercriminals to gather publicly available information about potential victims. By linking accounts, relationships, images, and videos at machine speed, these tools can make fraud and social engineering more convincing and scalable, lowering the barrier to entry for attackers. ### Source excerpt It's getting cheaper and easier for cybercriminals to research potential victims. Here's what's still in your control. ## Identity Abuse Through Trusted Communication Channels DevFeed: [Identity Abuse Through Trusted Communication Channels]() Original publisher: [Read original article]() Author: Bill Batchelor Published: 2026-08-20T10:00:25Z Content type: article Language: en Sources: [Unit 42]() Topics: [Security](), [Authentication](), [Social engineering](), [Software as a service](), [Incident response]() Tags: [authentication](), [credential-theft](), [identity](), [identity-theft](), [incident-response](), [malware](), [mfa](), [phishing](), [remote-access-software](), [saas](), [security](), [social-engineering](), [threat-research]() ### AI overview Unit 42 examines how attackers abuse trusted enterprise communication and collaboration platforms for identity phishing, impersonation, credential theft, malware delivery and social engineering. The article describes how compromised identities can make malicious activity appear legitimate within authenticated collaboration sessions and offers recommendations for detecting and defending against these attacks. ### Source excerpt Unit 42 details how attackers exploit enterprise collaboration tools for identity phishing and credential theft. Discover key defense strategies. The post Identity Abuse Through Trusted Communication Channels appeared first on Unit 42. ## How We're Building Scam Alert on WhatsApp With End-to-End Encryption and Verifiability Guarantees DevFeed: [How We're Building Scam Alert on WhatsApp With End-to-End Encryption and Verifiability Guarantees]() Original publisher: [Read original article]() Author: Chris Wiltz Published: 2026-08-12T13:00:28Z Content type: article Language: en Sources: [Engineering at Meta]() Topics: [End-to-End Encryption](), [Encryption](), [Machine Learning & Artificial Intelligence](), [Inference](), [Security](), [Bug Bounty](), [Social engineering]() Tags: [bug-bounty](), [encryption](), [machine-learning](), [privacy](), [scam](), [security](), [security-privacy](), [social-engineering](), [user-control](), [whatsapp]() ### AI overview An early technical overview of WhatsApp's optional Scam Alert feature, which uses a small on-device machine learning model to classify potential scam messages while keeping message content on the device. The article explains how the design preserves end-to-end encryption through local processing, avoids automatic reporting, gives users control, and supports independent security review during a limited Beta rollout. ### Source excerpt WhatsApp is committed to helping people stay safe while protecting the privacy of their messages. As scam tactics evolve -- from impersonation to social engineering to AI-generated lures -- we're always evolving as well, so that our protections stay ahead of scammers while protecting people's personal messages with end-to-end encryption. Today, we're sharing an early [...] Read More... The post How We're Building Scam Alert on WhatsApp With End-to-End Encryption and Verifiability Guarantees appeared first on Engineering at Meta. ## Inside the Modern SOC: The Identity Front Door DevFeed: [Inside the Modern SOC: The Identity Front Door]() Original publisher: [Read original article]() Author: Sharon Maydar Published: 2026-08-07T23:00:01Z Content type: article Language: en Sources: [Unit 42]() Topics: [Incident response](), [MFA](), [Vulnerabilities]() Tags: [ai](), [attacks](), [authentication](), [credential-theft](), [identity](), [incident-response](), [inside-the-modern-soc](), [insights](), [mfa](), [phishing](), [soc](), [social-engineering](), [unit-42-incident-response-report](), [vulnerabilities]() ### AI overview The article examines identity-based initial access, including credential theft, MFA manipulation, session hijacking, phishing, and social engineering. It describes how attackers expand access through persistence, privilege escalation, and lateral movement, challenging SOC teams to correlate signals across environments. ### Source excerpt Identity-based attacks drive 90% of incidents. Learn how modern attackers exploit identities and what SOC leaders can do to respond. The post Inside the Modern SOC: The Identity Front Door appeared first on Unit 42. ## Before the first prompt: Code execution paths in trusted coding-agent projects DevFeed: [Before the first prompt: Code execution paths in trusted coding-agent projects]() Original publisher: [Read original article]() Author: Nick Frichette Published: 2026-08-03T00:00:00Z Content type: article Language: en Sources: [Datadog Security Labs]() Topics: [AI-assisted coding](), [Claude Code](), [codex](), [Model Context Protocol (MCP)](), [Social engineering](), [Visual Studio Code](), [npm packages]() Tags: [agent](), [attacks](), [claude-code](), [code](), [codex](), [debugging](), [developers](), [mcp](), [npm-packages](), [social-engineering]() ### AI overview This security article examines how trusted coding-agent projects can execute repository-controlled code before a user sends the first prompt. It describes execution paths involving project-scoped MCP configuration in Codex and project-controlled environment settings and Git probes in Claude Code, without requiring a model response or shell-command approval. ### Source excerpt Learn how trusted coding-agent projects can execute repository-controlled code before the first prompt through Codex MCP configuration and Claude Code environment settings. ## ESET Threat Report H1 2026 DevFeed: [ESET Threat Report H1 2026]() Original publisher: [Read original article]() Author: Jiří Kropáč Published: 2026-07-08T08:45:00Z Content type: article Language: en Sources: [WeLiveSecurity]() Topics: [ESET research](), [Threat Research](), [Artificial Intelligence](), [Generative AI](), [Malware](), [Android](), [ClickFix](), [QR Code](), [ransomware](), [Social engineering](), [Endpoint Security & XDR](), [Security]() Tags: [ai](), [android](), [clickfix](), [code](), [eset-research](), [generative-ai](), [malware](), [ransomware](), [research](), [security](), [social-engineering](), [threat-report]() ### AI overview ESET's H1 2026 threat report describes attackers adapting established techniques across new platforms and behaviors. It highlights the expanding abuse of AI skills, PromptSpy Android malware using Google Gemini, the spread of ClickFix and QR-code phishing, and continued ransomware activity involving EDR killers. ### Source excerpt A view of the H1 2026 threat landscape as seen by ESET telemetry and from the perspective of ESET threat detection and research experts. ## AI coding tools can increase dependency and npm supply-chain risks DevFeed: [AI coding tools can increase dependency and npm supply-chain risks]() Original publisher: [Read original article]() Author: Shibam Dhar Published: 2026-06-22T00:00:00Z Content type: article Language: en Sources: [Harness Blog]() Topics: [ai-coding](), [Security](), [npm](), [Open Source]() Tags: [ai](), [ai-coding](), [npm](), [open-source](), [security](), [social-engineering](), [supply-chain-attacks]() ### AI overview The article explains that AI coding assistants can rapidly introduce vulnerable, malicious, or non-compliant open-source dependencies. It presents Harness Artifact Registry's Dependency Firewall as a registry-level control point for evaluating and blocking risky packages before they enter a CI/CD pipeline. ### Source excerpt AI coding tools can introduce vulnerable dependencies fast. Learn how dependency firewalls block risky packages at the registry level. Explore now. | Blog ## BTMOB: A stealthy RAT burrowing deep into Android devices DevFeed: [BTMOB: A stealthy RAT burrowing deep into Android devices]() Original publisher: [Read original article]() Author: Daniel Cunha Barbosa Published: 2026-05-26T08:50:00Z Content type: article Language: en Sources: [WeLiveSecurity]() Topics: [ransomware](), [Remote Access Trojan](), [Android](), [Social engineering](), [APK](), [Tooling](), [Accessibility]() Tags: [android](), [code](), [malware](), [phishing](), [remote-access-trojan](), [social-engineering](), [tool]() ### AI overview BTMOB is an Android remote access trojan that spreads through phishing websites, fake app stores, and malicious APKs. It can exfiltrate sensitive data, capture screenshots, record device activity, and enable remote control. Its APK builder and malware-as-a-service model make customized campaigns easier to launch. ### Source excerpt The malware pairs remote access capabilities with ready-made campaign tools, lowering the barrier for full device compromise ## What's New in Android Security and Privacy in 2026 DevFeed: [What's New in Android Security and Privacy in 2026]() Original publisher: [Read original article]() Author: Eugene Liderman Published: 2026-05-12T17:00:00Z Content type: article Language: en Sources: [Security]() Topics: [Android](), [Android Security](), [Security](), [spoofing](), [Social engineering](), [threat detection](), [On-device AI](), [Chrome]() Tags: [ai](), [android](), [android-security](), [fraud](), [none](), [on-device-ai](), [security](), [social-engineering](), [spoofing](), [threat-detection]() ### AI overview The article describes planned Android security and privacy enhancements for 2026, including verified financial calls to combat spoofed banking scams. Android can verify incoming calls through participating financial apps and automatically end calls that are not genuine. It also highlights expanded Live Threat Detection, which uses on-device AI to analyze app behavior and warn about suspicious activity. ### Source excerpt New Android security and privacy features ## Cyber resiliency in practice: Lessons from recent supply chain attacks DevFeed: [Cyber resiliency in practice: Lessons from recent supply chain attacks]() Original publisher: [Read original article]() Published: 2026-05-08T00:00:00Z Content type: article Language: en Sources: [Chainguard: Unchained]() Topics: [resiliency](), [supply chain attacks](), [supply-chain-security](), [axios](), [GitHub Actions](), [litellm](), [trivy](), [npm](), [PyPI](), [Python](), [Large Language Model]() Tags: [axios](), [chainguard-libraries](), [cyber-resiliency](), [cybersecurity](), [github-actions](), [litellm](), [malware](), [npm](), [pypi](), [python](), [security](), [security-attacks](), [social-engineering](), [software-supply-chain-security](), [supply-chain-attacks](), [trivy](), [worm]() ### AI overview The article defines cyber resiliency as an integrated ability to anticipate, withstand, recover from, and adapt to attacks across the full Protect, Detect, Respond, and Recover lifecycle. It examines recent supply chain attacks involving Trivy, Axios, and LiteLLM, highlighting risks such as GitHub Actions misconfiguration, long-lived tokens, incomplete credential rotation, mutable Git tags, maintainer-account compromise, and malicious package releases. ### Source excerpt Learn how modern cyber resiliency helps organizations prevent, detect, and recover from supply chain attacks like Trivy, Axios, and LiteLLM. ## New NGate variant hides in a trojanized NFC payment app DevFeed: [New NGate variant hides in a trojanized NFC payment app]() Original publisher: [Read original article]() Author: Lukas Stefanko Published: 2026-04-21T08:55:00Z Content type: article Language: en Sources: [WeLiveSecurity]() Topics: [Malware](), [Android](), [ESET research](), [genai](), [Social engineering](), [App](), [servers]() Tags: [android](), [attacks](), [brazil](), [eset-research](), [genai](), [malware](), [payment](), [social-engineering](), [threat-report]() ### AI overview ESET Research reports a new NGate malware variant hidden in a trojanized Android NFC payment app called HandyPay. The malware relays payment-card NFC data, steals card PINs, and exfiltrates them to an operator-controlled server. The active campaign, targeting users in Brazil since around November 2025, distributes the app through fake lottery and Google Play websites; the code may have been assisted by GenAI. ### Source excerpt ESET researchers discover another iteration of NGate malware, this time possibly developed with the assistance of AI ## As breakout time accelerates, prevention-first cybersecurity takes center stage DevFeed: [As breakout time accelerates, prevention-first cybersecurity takes center stage]() Original publisher: [Read original article]() Author: Phil Muncaster Published: 2026-04-07T09:00:00Z Content type: article Language: en Sources: [WeLiveSecurity]() Topics: [Cybersecurity](), [Threat Research](), [Artificial Intelligence](), [Automation](), [MFA](), [passwords](), [Reconnaissance](), [Social engineering](), [spoofing](), [ransomware]() Tags: [ai](), [ai-automation](), [attacks](), [authentication](), [automation](), [business-security](), [cybersecurity](), [exploits](), [malware](), [mfa](), [passwords](), [phishing](), [ransomware](), [social-engineering](), [tools]() ### AI overview Threat actors are using AI, automation, credential theft, phishing, zero-day exploits, reconnaissance, and AI-powered scripts to accelerate attacks. The article argues that shrinking breakout times require defenders to adopt a prevention-first cybersecurity strategy. ### Source excerpt Threat actors are using AI to supercharge tried-and-tested TTPs. When attacks move this fast, cyber-defenders need to rethink their own strategy. ## Designing AI agents to resist prompt injection DevFeed: [Designing AI agents to resist prompt injection]() Original publisher: [Read original article]() Published: 2026-03-11T11:30:00Z Content type: article Language: en Sources: [OpenAI News]() Topics: [AI Bots]() Tags: [agents](), [ai-security](), [attacks](), [chatgpt](), [security](), [social-engineering]() ### AI overview The article examines prompt-injection attacks against AI agents, arguing that effective defenses should constrain the impact of manipulation rather than rely solely on detecting malicious inputs. ### Source excerpt How ChatGPT defends against prompt injection and social engineering by constraining risky actions and protecting sensitive data in agent workflows. [Next page]()