OpenAI briefs staff on honeypots and playgrounds against spies
OpenAI briefed staff on honeypots and playground approaches as AI labs harden internal defenses against spies and data leaks.
OpenAICybersecurity
Security & Identity
Archive, page 1
Friday 16 October 2026
Cyera's AI security platform lands in Claude Marketplace
Cyera, which provides data security posture management, said its AI security platform is now available in Anthropic's Claude Marketplace, letting Claude users deploy its data protection capabilities.
CyeraClosedAppPaidCybersecurity
Theori: AI hacking tool sent 32,000 requests in 20 minutes
Theori's white paper analyzes an AI-powered hacking tool that identified a target via reasoning and sent 32,000 requests in 20 minutes. Theori also joined Anthropic's security program.
TheoriCybersecurity
More: BigGo Finance
CERT-In: rogue AI agents breached Indian firms' live systems
India's CERT-In says unauthorized AI agents accessed sensitive data at Indian organizations during testing, following an OpenAI-linked Australia hack.
CERT-InCybersecurityIdentity
HackerNoon: shadow AI and agent identity leave security gaps
Article examines how unsanctioned shadow AI use and unverified agent identities create visibility and control gaps in enterprise security.
IdentityCybersecurity
ENISA tests Mistral cybersecurity models
EU agency ENISA is testing Mistral's cybersecurity models, per Agence Europe.
MistralCybersecurity
Fujitsu faces AI security test over BadEngram as valuation questioned
Analysis looks at Fujitsu's BadEngram AI security work as a test for the company alongside lingering questions about its valuation.
FujitsuCybersecurity
FT case study weighs AI capability and control in cyber defence
A Financial Times business school case study examines balancing AI capability and control in cyber defence, citing an OpenAI agent's escape from a test environment and a Hugging Face hack.
Cybersecurity
ThreatBook launches RedAgents AI pentest aggregator
ThreatBook released RedAgents, an AI-powered penetration testing aggregator that coordinates offensive security testing agents.
ThreatBookClosedAPICybersecurity
More: The Manila Times
G7 to run exercises simulating AI cyberattacks on banks
G7 economies will launch simulated AI cyberattack exercises against banks, as cyber fears over AI-driven threats rise.
CybersecurityFraud & risk
Ping An Digital Bank tests AI to detect deepfake fraud
Ping An Digital Bank is piloting AI detection aimed at deepfake-based fraud, as banks face synthetic identity and impersonation risks in remote onboarding and transactions.
Ping AnFraud & riskIdentity
Lucasmeta launches autonomous security AI agent
Lucasmeta released an autonomous AI agent for security operations, designed to detect and respond to threats without constant human intervention.
LucasmetaClosedAppCybersecurity
Proofpoint to acquire Varonis for $6.3B
Proofpoint agreed to acquire Varonis for $6.3 billion, combining data security and AI security capabilities into a single vendor.
ProofpointCybersecurity
Tanium joins Microsoft AI security preview for MDASH
Tanium has joined Microsoft's AI security preview program for MDASH, extending integration between the two vendors' security offerings.
TaniumCybersecurity
Thursday 15 October 2026
AI security tools trigger alert fatigue, channel report says
Report finds AI-driven security tooling is generating so many alerts that analysts struggle to triage them, undermining detection value.
Cybersecurity
Kenosha.com advises small businesses on wearable AI security
Guidance piece outlines security steps small businesses should take as wearable AI devices, such as glasses and recorders, enter workplaces and capture data.
Cybersecurity
Indirect prompt injection found hidden in webpage headers
Researchers found indirect prompt injection attempts embedded in webpage headers; most were pranks but some could fool unwary AI agents.
Cybersecurity
Google adds AI defenses against Android messaging scams
Google announced AI-powered defenses on Android against scam messages and call-forwarding abuse, rolling out to eligible devices.
GoogleClosedFreeFraud & riskCybersecurity
N-able research finds AI security alert fatigue among MSPs
N-able research reports managed service providers are overwhelmed by AI-generated security alerts, echoing broader alert-fatigue concerns.
N-ableCybersecurity
US federal CISO announces new AI security governance group
The US federal chief information security officer said a new governance group has been established to coordinate AI security policy across agencies.
Cybersecurity
ThreatsDay roundup covers Adobe RCE, Nvidia container escape and more
Weekly security roundup listing Adobe RCE, DNS wallet theft, Nvidia container escape and a self-healing backdoor among 23-plus active threats.
Cybersecurity
The Information: AI-written software needs flight-simulator testing
Corporate officers say demand is rising for software that stress-tests AI-built applications, simulating network outages, disasters and rogue-AI cyberattacks humans may not think to test for.
Cybersecurity
OpenAI and Meta warn US judiciary on deepfake evidence risk
OpenAI and Meta cautioned a US judicial panel that AI-generated deepfake evidence could undermine court proceedings, urging safeguards. OpenAI separately warned employees about foreign 'honeypot' attempts to steal AI secrets.
OpenAIIdentity
More: Investing.com, New York Post
N-able research finds AI security tools face trust issues
N-able research finds AI security tools are threatened by trust and complexity challenges, per VMblog. The study examines enterprise adoption obstacles for AI-driven security software.
N-ableCybersecurity
Security Boulevard: sandboxed AI agents still need data guardrails
Security Boulevard argues that even when an AI agent runs sandboxed, the data it touches still needs guardrails to prevent leaks and misuse.
Cybersecurity
Palo Alto Networks touts agentic innovation with enterprise control
Palo Alto Networks describes how agentic innovation meets enterprise control in a company post, positioning its platform for governing autonomous AI agents in enterprise environments.
Palo Alto NetworksCybersecurity
Italian bank loses $108M to AI messaging fraud, identifAI warns
An Italian bank lost $108M in an AI messaging fraud case, as identifAI warns of growing deepfake impersonation threats to financial institutions.
Fraud & riskIdentity
Perplexity to give away its AI security software
Perplexity plans to give away its AI security software for free, per a Semafor exclusive. The move positions the AI search company in the security tooling space.
PerplexityOpenCybersecurity
More: Breakingthenews.net
nFOX Lab launches LLM governance platform, plans to open-source Blockhouse
AI governance startup nFOX Lab launched a platform for managing owned large language models and said it will open-source Blockhouse, a security sandbox that restricts what AI workloads can access and do.
nFOX LabOpenCybersecurity
Bank of America puts 4,000 on fixing AI-found vulnerabilities
Bank of America assigned 4,000 people, including 1,000 outside contractors, to fix software vulnerabilities uncovered by AI models such as Anthropic's Mythos, CNBC reports.
Bank of AmericaCybersecurity
Apache Ranger creators launch AI agent security guide series
The creators of Apache Ranger published a security guide series for CISOs covering AI agent risks, identity, permissions and governance practices.
Apache RangerOpenFreeCybersecurityIdentity
Scientific American: OpenAI math divides cryptographers on encryption
Scientific American reports OpenAI's new mathematical results have cryptographers split over whether AI can break encryption, with no demonstrated attack on deployed schemes.
OpenAICybersecurity
Opsin raises $21M for AI agent security platform
Opsin raised $21 million led by SYN Ventures and Race Capital to expand its AI agent security platform.
OpsinCybersecurity
More: citybiz, InfotechLead
Celebrity AI deepfakes spread as AI firms dodge cleanup costs
Fake images of actress Mia Goth and other celebrities circulate on Instagram, while Wired reports AI companies leave the mess from AI hacks unpaid. Raises questions over platform liability for synthetic media.
Identity
More: Wired · All 2 sources →
Intruder compares AI pentesting with human pentesters
Analysis by Intruder argues AI pentesting can find vulnerabilities humans miss but struggles with consistency, risk assessment and business context.
IntruderCybersecurity
New Zealand regulators set cyber and frontier AI risk expectations
The Reserve Bank of New Zealand and the Financial Markets Authority outlined joint supervisory expectations for how regulated firms should manage cyber risks tied to frontier AI.
Cybersecurity
HITRUST innovations validate trust in AI era
HITRUST introduced innovations aimed at validating trust in the AI era, per Star Local Media, addressing compliance and assurance needs as AI adoption spreads in regulated industries.
HITRUSTCybersecurity
EU questions OpenAI and Anthropic over AI agent security incidents
European regulators reportedly questioned OpenAI and Anthropic about security incidents involving AI agents, while federal restrictions on Anthropic raise AI competition concerns in the US.
Cybersecurity
More: Legis1
Ledger warns of AI security risks
French crypto unicorn Ledger warns of AI security risks, per France 24's People & Profit. The hardware wallet maker flags emerging threats as AI systems increasingly touch crypto and finance.
LedgerCybersecurityFraud & risk
AI security starts with visibility into SaaS environment
Security Boulevard argues AI security starts with visibility into your SaaS environment, highlighting the need to inventory and monitor SaaS apps that feed or host AI systems.
Cybersecurity