COGNOSCERE Daily Tech Review — Issue T134 · Tuesday, June 2, 2026

Or visit Intelligence Overview for deeper analysis.

ACTPalo Alto GlobalProtect VPN Auth Bypass Flaw (CVE-2026-0257) Now Exploited in Attacks

Attackers are actively exploiting CVE-2026-0257, an authentication bypass vulnerability in Palo Alto Networks PAN-OS GlobalProtect that enables unauthorized VPN sessions via forged override cookies. CISA added the flaw to its Known Exploited Vulnerabilities catalog with a June 1 federal agency remediation deadline, as Rapid7 confirmed successful exploitation across multiple enterprise customers beginning May 17.

BleepingComputer · Cybersecurity · Relevance: 1.0 · Source →

VPN, authentication-bypass, active-exploit, CISA-KEV, network-security, patch-management, zero-trust

ACTHackers Exploit FortiClient EMS Flaw to Push EKZ Infostealer Malware

Threat actors are exploiting CVE-2026-35616 (CVSS 9.1), a critical unauthenticated access-control bypass in FortiClient Enterprise Management Server, to push a previously undocumented credential stealer called EKZ disguised as a legitimate Fortinet endpoint patch. Arctic Wolf observed the campaign abusing FortiClient's own VPN scripting workflows to silently harvest browser credentials, cookies, and payment data from all managed enterprise endpoints.

BleepingComputer · Cybersecurity · Relevance: 0.9 · Source →

endpoint-security, credential-theft, infostealer, malware, enterprise-management, active-exploit, patch-management

ACTFake Anthropic Sites Deliver Fileless Infostealer to Claude Code Users via ClickFix Campaign

Cyderes uncovered an active SEO-poisoning and ClickFix campaign spoofing Anthropic download pages to trick Claude Code users into running a malicious mshta.exe command that reflectively loads a .NET infostealer entirely in memory, beaconing browser credentials to Russian infrastructure. The attack chain exploits enterprise AI tool adoption as a lure, with no files dropped on disk, making endpoint detection challenging.

TLDR InfoSec · Cybersecurity · Relevance: 0.9 · Source →

social-engineering, ClickFix, infostealer, SEO-poisoning, fileless-malware, AI-tool-targeting, credential-theft

PREPARECisco Live 2026 Las Vegas Pushes Enterprise Networking Toward AgenticOps and AI-First Automation

Cisco opened its annual Las Vegas conference (May 31–June 4, 20,000 attendees) centered on its 'AgenticOps' vision, in which AI agents autonomously handle routine IT tasks including troubleshooting and provisioning. CEO Chuck Robbins is headlining keynotes as Cisco pursues a $9 billion AI infrastructure order target, with announcements spanning AI-driven security operations, agentic network automation, and expanded AI Defense capabilities.

TechTimes · Enterprise Networking · Relevance: 0.8 · Source →

network-automation, agentic-AI, AI-infrastructure, enterprise-networking, AIOps, zero-trust, Cisco-Live-2026

WATCHWipro Expands ServiceNow Partnership to Embed Agentic AI Workflows Across Core Enterprise Functions

Wipro announced an expanded integration of its Wipro Intelligence platform with the ServiceNow AI Platform to deploy governed agentic AI workflows across IT, HR, procurement, and cybersecurity operations at enterprise scale. The partnership includes named solutions such as SmartProcure and Cyber Transform, and triggered a roughly 4% intraday share price gain for Wipro on announcement day.

Wipro Newsroom · Enterprise Software / SaaS · Relevance: 0.8 · Source →

agentic-AI, enterprise-automation, IT-operations, HR-tech, procurement, cybersecurity-operations, partnership

WATCHDutch Authorities Dismantle 17-Million-Device Botnet Linked to ASOCKS Proxy Service

Dutch police and the National Cyber Security Centre seized 200 servers and dismantled one of the largest residential proxy botnets ever disrupted, comprising at least 17 million compromised consumer devices worldwide. The infrastructure, linked to Russia-based ASOCKS, routed cybercriminal traffic through ordinary consumer hardware to defeat IP-reputation defenses, with devices enrolled via malware without owner consent.

BleepingComputer · Cybersecurity · Relevance: 0.8 · Source →

botnet, law-enforcement-takedown, residential-proxy, threat-infrastructure, IoT-security, cybercrime

PREPAREGartner Warns: Uniform AI Agent Governance Will Cause 40% of Enterprise Agent Deployments to Fail by 2027

Gartner issued a formal research warning that enterprises applying one-size-fits-all governance to AI agents—regardless of autonomy level—will encounter systematic failure, and predicted 40% of enterprises will demote or decommission autonomous agents by 2027 due to governance gaps discovered only after production incidents. The firm recommends a proportional four-tier framework (Observe, Advise, Act-with-Approval, Fully Autonomous) with governance controls scaled to each tier's trust boundary and action scope.

Gartner · AI Safety and Alignment / Governance · Relevance: 0.9 · Source →

AI-governance, agentic-AI, risk-management, enterprise-AI, AI-safety, compliance, C-suite

PREPAREMicrosoft Copilot Super App Revealed: GitHub Copilot, Cowork, and Always-On Scout Agent Unifying by Summer 2026

Microsoft is building a unified 'Copilot super app' under the internal slogan 'Delivering one Copilot,' combining GitHub Copilot, Copilot chat, Cowork, and an agentic workflow layer called Autopilot into a single desktop interface targeting a late-summer 2026 launch. Leaked screenshots surfaced ahead of Microsoft Build (June 2–3, San Francisco) showing a Scout always-on agent and personal-versus-enterprise account toggling, with CEO Satya Nadella expected to unveil new proprietary AI models at the event.

Fortune · Agentic Systems · Relevance: 0.9 · Source →

AI-super-app, agentic-AI, enterprise-AI, Microsoft-Build-2026, productivity-software, GitHub-Copilot, AI-agents, SaaS

PREPAREMicrosoft Says 2026 Is the Year AI Agents Become Operational Workers Across Microsoft 365

Microsoft announced a strategic positioning shift framing AI agents as operational workers rather than productivity assistants, signaling an intent to embed workflow execution—not just augmentation—across Microsoft 365 environments. The announcement aligns with broader Build 2026 messaging and reinforces enterprise demand for agent governance, access controls, and audit capabilities within existing Microsoft tooling.

Microsoft Partner Center · Enterprise AI Adoption · Relevance: 0.8 · Source →

enterprise-AI, agentic-AI, Microsoft-365, workflow-automation, AI-governance, future-of-work

WATCHMiniMax M3 Launches: First Open-Weight Model to Combine Frontier Coding, 1M-Token Context, and Native Multimodality

Shanghai-based MiniMax officially released M3 on June 1, 2026, an open-weights multimodal model claiming 59.0% on SWE-Bench Pro (ahead of GPT-5.5's 58.6%) with a 1-million-token context window powered by a new MiniMax Sparse Attention architecture at roughly 1/12th the API cost of closed frontier models. Open weights and the technical report are expected on Hugging Face within ten days; benchmarks are currently vendor-reported and await independent verification.

TechTimes · AI Model Releases · Relevance: 0.9 · Source →

open-weights, LLM, agentic-coding, multimodal, Chinese-AI, frontier-model, cost-efficiency, data-sovereignty

WATCHHyland Unveils Enterprise Agent Mesh and AI Control Tower Capabilities for Governed Agentic AI at Scale

Hyland announced general availability of its Enterprise Context Engine and the Enterprise Agent Mesh, which provides governed orchestration and lifecycle management for AI agent fleets across healthcare, banking, insurance, and government sectors. The release also introduced Agent Lifecycle Management and a Control Tower with an Agent Library catalog to prevent agent sprawl, alongside headless API mode enabling integration with Databricks and Snowflake ecosystems.

Help Net Security · Enterprise AI Adoption · Relevance: 0.8 · Source →

agentic-AI, AI-governance, enterprise-content-management, agent-orchestration, regulated-industries, AI-platform

PREPAREDatabricks 2026 State of AI Agents Report: Governance Tools Drive 12x More Projects to Production

Databricks' State of AI Agents report, drawing on data from more than 20,000 global customers, found that AI governance and security products saw the largest usage increase of any category and that companies using governance tools moved 12 times more AI projects into production than those without. The report also highlights a rapid shift from chatbot architectures to multi-agent systems and an associated drive toward Lakebase-style transactional database transformation.

Databricks Blog · Decision Frameworks / Risk Assessment · Relevance: 0.8 · Source →

agentic-AI, AI-governance, enterprise-AI, multi-agent, data-strategy, ROI, decision-support

WATCHDatabricks Lakebase Introduces Git-Style Copy-on-Write Database Branching for AI Agent and Developer Workflows

Databricks added copy-on-write database branching to its Lakebase managed Postgres service, enabling developers and AI agents to spin up isolated, production-scale database clones in one second at zero initial storage cost—solving the long-standing problem of shared staging environments. The capability integrates with CI/CD pipelines and is also available as a programmable primitive for AI agents requiring ephemeral, governed data environments.

Databricks Blog · Data Infrastructure / Analytics Tools · Relevance: 0.8 · Source →

data-infrastructure, database, PostgreSQL, AI-agents, developer-productivity, data-governance, DevOps, analytics