Latest coverage for Artificial Intelligence

Strategic Plan Includes Human Risk Management Platform Expansion, IPO PreparationBryan Palma outlines his vision to grow KnowBe4 beyond security awareness training by investing in agentic AI, expanding email and behavioral tools and positioning the company for IPO readiness. He highlights Vista Equity's support and platform depth as key assets.

Also: Cyber IPOs and the Investment Climate, the Urgency of AI ExplainabilityIn this week's update, ISMG editors unpacked Trump's teased "grand cyber plan" amid budget cuts to the Cybersecurity and Infrastructure Security Agency, key business takeaways from RSAC Conference 2025 and why explainability in artificial intelligence is becoming critical to trust and security.

Partners Use Bedrock, SageMaker for Threat Detection, Response, Vital to InnovationAWS is enabling cybersecurity firms to enhance detection, triage and response capabilities by embedding generative AI into services like Bedrock and SageMaker, while reinforcing its position as a partner-centric cloud security leader, said Managing Director Rohan Karmarkar.

University of Texas CISO George Finney on Zero Trust Challenges and His New BookEnterprises need to mature their zero trust models to recognize how trust is inherently built into artificial intelligence and how to proactively identify vulnerabilities. George Finney, CISO at University of Texas Systems, says security teams need to be trained to spot implicit trust across systems.

Old Visuals, AI Deepfakes and Fake Claims Go Viral Amid Escalating ConflictMisinformation is going viral in India as the New Delhi government called Friday on the financial sector to strengthen cyber defenses amid growing military activity along the Pakistani border. Tensions between the two countries ratcheted significantly upward Friday.

Pixtral Models 60 Times More Likely to Generate Harmful Content Than RivalsPublicly available artificial intelligence models made by Mistral produce child sexual abuse material and instructions for chemical weapons manufacturing at rates far exceeding those of competing systems, found researchers from Enkrypt AI.

CISOs Seek Real Value as Vendors Tout the Latest Batch of AI-Driven SolutionsAs the conversation shifts from generative to agentic AI, it's clear that AI holds tremendous potential to ease zero trust fatigue, but only when guided by business context, quality data and human oversight. CISOs see AI as a "basket of opportunities but plenty of "vendor blind spots."

Funding Will Fuel R&D Push Into Automated Remediation and Risk Prioritization ToolsWith code increasingly generated by AI and attackers using AI for exploits, OX Security raised $60 million to scale R&D and help developers prioritize critical vulnerabilities. The company aims to close detection gaps and reduce time-to-remediation in application security.

Agentic AI Is More Than Just the Latest Cybersecurity Buzzword at RSAC ConferenceAt RSAC 2025, the message came through loud and clear: Agentic AI is no longer just a concept. It's being deployed today. While much of the buzz focused on performance gains and trust concerns, another story emerged - one that speaks directly to security professionals and those entering the field.

California Man Pleads Guilty to Two Felony Charges Related to Hacking Employee's PCA California man agreed to plead guilty to hacking a Disney employee's personal computer and stealing over one terabyte of confidential company data. Authorities say the man posted a malicious artificial intelligence art application online and used it to steal an employee's credentials.

NCSC Expects Attack Volume by 'Advanced Threat Actors' to Rise Sharply by 2027Proliferation of AI-enabled technology will widen access to offensive tools by nation-state groups and other hackers. The volume of attacks is expected to rise significantly by 2027, and British critical infrastructure will be a prime target, the National Cybersecurity Center said.

Loss of 5% of Staff Is Cybersecurity Industry's Second-Largest Workforce ReductionCrowdStrike plans to ax 500 employees as the endpoint security behemoth looks to operate more efficiently and uses artificial intelligence to "flatten[s] our hiring curve. The company revealed plans to reduce its nearly 10,000-person staff by 5% to scale its business with more focus and discipline.

New Research Uncovers Tokenizer Blind Spots in Leading LLMsSubtle obfuscation techniques can systematically evade the guardrails that today's large language models rely on. Researchers from Mindgard team found that adversaries can "smuggle" malicious payloads past tokenizers using emojis, zero-width spaces and homoglyphs.

Chris Wysopal of Veracode on How AI Boosts Code Production and VulnerabilitiesAI delivers a 50% increase in developer productivity, but with more code comes more vulnerabilities. Chris Wysopal, chief security evangelist at Veracode, shares developments in secure code practices and how regulatory pressures are improving prioritization of secure code.

Ex-Deputy NSA Anne Neuberger on Preparing for AI-Driven ThreatsAnne Neuberger, former deputy national security advisor for cyber and emerging technologies, White House, outlines the urgent need for resilient critical infrastructure, strategic AI use in cybersecurity, and enhanced federal-state coordination to protect against evolving cyberthreats.

Michelle Dennedy of Abaxx on Why Privacy Demands Curiosity and AccountabilityAI has reignited a debate on privacy. While some argue the era of personal data protections is over, others see AI as an opportunity to apply precision, nuance and contextual decision-making to data governance, said Michelle Dennedy, chief data strategy officer, Abaxx Technologies.

Glilot Capital Partners' Kleinstein on How Gen AI Transforms the Security LandscapeArik Kleinstein, co-founder and managing partner, Glilot Capital Partners, says startups have an advantage over incumbents because they don't have to deal with legacy technology. But he shared some steps startups can take to secure their data and AI models.

'Building Fast' While 'Building Competently' Remains Key, Expert SaysQuestions over the risk posed by artificial intelligence dominated discussions at this year's RSAC Conference in San Francisco. Experts said that building models - at speed - that are secure and reliable remains essential for creating AI tools organizations will trust and want to adopt.

Trellix's John Fokker Advises CISOs to Prioritize Patching, MFA, Network VisibilityThreat actors aren't rushing to adopt AI tools to exploit vulnerabilities. "They still prefer a victim with weak passwords, bad MFA, bad patching. It is the easiest way to make money for criminals so they don't have to invest in AI," said John Fokker, head of threat intelligence at Trellix.

HackerOne CEO Warns AI Can't Replace Creativity, Intuition in Cyber Defense EffortsDespite AI advances, human intuition remains essential in security, says HackerOne CEO Kara Sprague. Machines detect patterns, but only people can anticipate the unexpected. Ethical hackers and human-led defense are vital to addressing evolving threats and the cybersecurity talent gap.