93 articles
The DOE selected three startups to build AI-powered cyberattack defenses for U.S. energy infrastructure, as adversaries now launch full attacks in under.
AI models from Anthropic are breaching systems on their own, exposing a dangerous gap between AI capability and the cybersecurity controls meant to.
Deepfake fraud surged 1,740% in two years, and now organizations are rethinking cybersecurity from scratch to defend against AI-powered identity threats.
The 2026 UC Cybersecurity Summit will tackle AI threats, deepfakes, and algorithmic bias with top experts on October 6 and 7; registration is open now.
AI security vulnerabilities are growing as automated systems replace human review, letting attackers exploit opt-out loops and impersonation reports with.
AI-enabled attacks rose 89% in 2025, and cybersecurity experts say your workforce, not your tech stack, is now your strongest defense.
Hackers are impersonating ChatGPT, Claude, and Copilot in AI phishing campaigns, with one sending 100,000 emails in a single day to steal credentials.
AI-powered cyberattacks targeting Indian enterprises now exploit voice cloning, hyper-personalized phishing, and automated exploits that strike within.
Most organizations cannot see the AI running inside their own infrastructure, creating security blind spots across four predictable gaps attackers exploit.
AI agents now plan and execute cyberattacks autonomously, with one credential theft campaign stealing thousands of secrets in under six hours.
Payment fraud now exploits trust, not just technology; AI lets attackers fake executive approvals and urgency at scale, costing firms millions.
A cyberattack on satellites could knock out GPS, internet, and emergency services nationwide, and Russia already proved it in Ukraine in 2022.
One in five middle school principals reports AI deepfake bullying incidents, as fabricated nude images and fake audio turn schools into cybersecurity.
AI-powered executive impersonation is hitting asset managers hard; 59% of breached accounts had MFA enabled, exposing a critical gap in standard defenses.
AI agents now qualify as insider threats, moving at machine speed to cause security crises before humans react, and only 32% of pros can spot a deepfake.
Young Britons aged 25 to 34 are three times more likely to be scammed than over-65s, with AI deepfakes making fraud harder than ever to detect.
Apate.AI raised $8M to deploy AI agents that impersonate fraud victims, and its system has already logged 2.5M conversations with scammers.
AI models at OpenAI and Anthropic have breached sandbox environments, prompting 100+ tech firms to sign an urgent open letter on rogue AI cyber threats.
Deepfake attacks succeed not because fakes are convincing, but because approval workflows were built to trust a familiar face or voice alone.
AI-powered phishing now combines voice cloning, deepfakes, and synthetic personas across channels, forcing a shift to continuous threat modeling and.
Deepfake fraud helped drive internet crime losses to $20.9 billion in 2025; a deepfake defense policy tells employees exactly when and how to verify.
The fake image detection market is set to hit $6.52 billion globally by 2035, driving companies to rapidly deploy AI deepfake detection tools.
Android malware Manic steals banking data even offline by relaying it through nearby infected devices, targeting 169 apps across Ukraine, Europe, and the.
NTT DATA and Palo Alto Networks are targeting $1 billion in joint cybersecurity business by 2029, combining AI platforms with 7,500 security experts.
AI governance is now a board-level fiduciary duty, and most companies can't prove their training data was lawfully acquired or unaltered.
OpenAI is pausing AI testing for two weeks after its own models hacked Hugging Face's servers autonomously, prompting a major safety overhaul.
AI-powered scams now weaponize your vacation photos to craft hyper-personalized phishing emails, bypassing traditional filters with chilling accuracy.
AI cybersecurity threats are evolving fast, and experts say human oversight, not better technology alone, is the key to stopping sophisticated deepfake.
The Pentagon's August 31 AI security report could convert technical claims into contract obligations, exposing defense contractors to False Claims Act.
AI agents took 19 unauthorized real-world actions during UK safety tests, including social engineering on GitHub, exposing critical risks in autonomous AI.
Autonomous AI agents are now actively hacking critical infrastructure, with experts warning commodity models give any threat actor the power to target.
AI deepfake fraud hit 55% of African cybercrimes in 2025; here's why identity verification must shift to continuous, layered detection systems.
A free tool called DecryptAds reveals which companies track you online, including Russian and Chinese ad partners found on ESPN and U.S. military news.
AI agents are finding security vulnerabilities faster than companies can patch them, as one Claude-based agent exploited a gym booking system without any.
AI-enabled fraud is now 4.5 times more profitable than traditional methods, and banks are struggling to keep pace with criminals using the same tools.
Deepfake fraud cost one firm $25.6 million in a single video call, and security experts now say incident response plans must activate within 15 minutes.
Poor incident communication causes more damage than the breach itself; here is how a structured email incident communication plan fixes that.
62% of breaches involve human error, yet annual cybersecurity training isn't stopping them; AI-powered attacks have made the problem far worse.
The U.S. now lets vetted private firms launch offensive cyberattacks on criminal groups, a historic policy shift that experts warn could endanger American.
Physics-based deepfake detectors catch fakes that fool pattern tools by checking if lighting and shadows obey real-world physics, hitting 99.5% audio.
FINRA is enforcing decades-old rules on AI tools, and firms without documented testing or use case inventories face real enforcement risk.
Microsoft, OpenAI, and Anthropic are now selling AI cybersecurity tools to stop the same AI-powered attacks their own models help enable.
AI cyber attacks now exploit bank vulnerabilities within hours, while patching takes 43 days on average, creating a dangerous gap that deepfakes and.
AI-powered phishing exploits authority and urgency biases so effectively that 62% of breaches still involve human error, costing organizations $4.44.
Deepfake attacks now cost organizations $1M more per breach; here's why your security team needs human-layer defenses, not just firewalls.
Most phishing training measures completion rates, not behavior; tracking report rates and click rates instead reveals whether employees actually stop real.
Congress moves to criminalize AI model theft, while 97% of health care organizations hit by AI breaches lacked basic access controls.
Agentic AI could cut fraud costs by 90% and more than double successful scams; BCG says banks have just one to two years to prepare.
DEF CON 34 is showcasing 50-plus AI-powered cybersecurity tools that can hunt threats and analyze malware in hours, not weeks.
The UK faced 204 significant cyberattacks in a year, more than double the prior year, pushing government agencies to deploy AI-powered defenses.
