Latest coverage for Artificial Intelligence

AI-Developed Attack Tooling Generated 'High-Volume, Noisy Workflows'A hacker used Claude and Chat GPT in a cyberattack against a municipal water and sewage utility's operational technology systems in Mexico in January, according to forensic analysis by OT security firm Dragos. The tools "leveraged known techniques and existing vulnerability knowledge."

Also: Washington's AI Policy Divide, FDA's Push for AI-Driven Clinical TrialsIn this week's panel, four ISMG editors discussed the battle over who gets to access powerful AI cybersecurity models, policy issues unfolding in Washington over AI-driven cyber defenses, and how the FDA is beginning to test AI-supported real-time clinical trials to speed up drug development.

Study Finds Standard Safety Tests Miss Most Agentic AI ThreatsResearchers from Stanford, MIT, Carnegie Mellon and others found that most production AI agents are vulnerable to attacks that unfold across multi-step actions. The study warns that memory, tool access and agent coordination create failure modes traditional chatbot safety testing cannot see.

Pentagon Expands Frontier AI Providers Amid Anthropic Legal FightThe Pentagon said it will no longer depend on a single artificial intelligence provider as the White House pushes agencies to diversify frontier AI systems amid an escalating legal and policy fight with Anthropic over military use of advanced models.

Artificial intelligence that is embedded in newer editions of software and other technology tools but is not explicitly revealed by vendors is a substantial risk on par with shadow AI, said regulatory attorney Elizabeth Hodge with the law firm Akerman LLP.

ServiceNow Takes Aim at Enterprise AI Sprawl at Knowledge 2026At its Knowledge 2026 conference, ServiceNow announced artificial intelligence control tower expansions, an autonomous workforce across every business function and a platform play to become the operating layer for all enterprise AI solutions.

Trilogue Deal Carves Out Industrial AI, Adds Nudifier BanLawmakers from Europe's political institutions agreed to water down the continent's landmark artificial intelligence regulation at a moment when the 2024 AI Act has barely started to be implemented. The law's requirements for high-risk AI will likely only be enforced starting in December 2027.

AI-Driven Attacks Compress Breakout Times, Forcing Defenders to Rely on Context NowAI has lowered the cost and speed of cyberattacks, enabling adversaries to exploit vulnerabilities within minutes. As breakout times collapse, security teams must respond faster by using context-driven intelligence and automation to detect, prioritize and stop threats in real time.

ARPA-H Program Aims to Speed Up Disease Breakthroughs Using AI-Enabled EcosystemBiomedical research breakthroughs for complex diseases and chronic illnesses can take years to achieve. The U.S. Department of Health and Human Services is hoping to speed that up ten-fold by creating an artificial intelligence-enabled interoperable research ecosystem.

Mythos Found ‘Tens of Thousands’ of Unpatched Flaws With Months to Fix ThemAnthropic CEO Dario Amodei warned that Claude Mythos has found tens of thousands of unpatched software vulnerabilities, with a six-to-12 month window before Chinese AI models catch up. The disclosure came alongside a major financial services push including an investor-backed firm and 10 new AI agents.

Mythos Moves the Needle on AI Innovation, DefenseAnthropic’s “Mythos moment” is accelerating vulnerability discovery, but speed without validation is a growing risk. As exploit windows shrink and remediation lags, more findings only mean more noise. The real advantage lies in validating what actually matters—and fixing it first.

Agent Promises to Cut Investigation Time from Days to MinutesFIS has partnered with Anthropic to deploy an AI agent that automates money laundering investigations, aiming to reduce casework from days to minutes. BMO and Amalgamated Bank are first in line, with broader availability targeted for late 2026.

BlueVoyant Seeks to Expand Beyond MDR Clients Into Firms With Mature In-House SOCsBlueVoyant named John Hernandez - the former leader of Quest's Microsoft security business - as its next CEO to drive an agentic AI SaaS platform that expands the vendor beyond managed services and helps customers accelerate detection, response and supply-chain risk management.

Benchmarks Shows Matched Capability, Brittle ReasoningTwo artificial intelligence models from competing labs have essentially the same offensive cyber capability level, with consistent reasoning failures that the cyber scores alone do not capture. OpenAI's GPT-5.5 and Anthropic's Mythos Preview now deliver near-identical offensive cyber performance.

Mythos a Turning Point, Say Lawmakers in Missive to European CommissionDozens of European lawmakers are pressing the European Commission to act quickly to protect the continent's cybersecurity, due to the advent of new AI models that have considerable hacking prowess.

Guidance Warns Autonomous Systems Expand Enterprise ExposureFederal and Five Eyes cyber agencies warn that agentic AI systems - capable of autonomous action across enterprise environments - are introducing identity, visibility and control risks that could outpace existing defenses without continuous monitoring, zero trust enforcement and human oversight.