LIVE · cybersecurity feed
Live wire
ai

Microsoft Copilot Deployments Delayed Over Security Concerns

CoreView research finds that security leadership is concerned about AI Assistant exposing confidential data

zeroday.news · 9d ago

A significant majority of organizations have either postponed or canceled their planned deployments of Microsoft Copilot, citing substantial security concerns regarding the AI-powered assistant's potential to expose sensitive internal data. A report published on July 21, 2026, by CoreView, titled "State of Microsoft 365 Security and Governance 2026," indicates that two-thirds of organizations are hesitant to proceed with Copilot's integration due to these worries.

The primary apprehension centers on the risk of Copilot surfacing confidential information, particularly from SharePoint. Many organizations reportedly struggle with data governance issues within SharePoint, leading to confusion about the access and permissions granted to Copilot. This lack of clarity fuels fears of data leakage or the inadvertent sharing of SharePoint links outside the organization.

The decision to delay or cancel deployments is most prevalent among senior leadership. The survey found that three-quarters of C-level executives have instructed their organizations to pause Copilot rollouts, a sentiment shared by 60% of managers. CoreView CEO Simon Azzopardi noted that executives are particularly attuned to the potential for AI to highlight long-standing issues with shared links and unmanaged permissions, making a previously latent risk both visible and urgent.

Of those organizations that have delayed or canceled Copilot deployment, 73% specifically attributed their decision to concerns that the AI could uncover confidential internal information. The report characterizes this widespread organizational hesitation as striking, especially for a flagship Microsoft product with considerable executive interest.

CoreView suggests that this reluctance often stems from prior cybersecurity incidents involving Microsoft 365. These incidents were frequently linked to the absence of fundamental security controls, such as multi-factor authentication (MFA) for administrators, privileged access management (PAM), or configuration tamper detection. Cybersecurity leaders are concerned that similar vulnerabilities could arise with Microsoft Copilot.

To mitigate these risks, experts recommend implementing robust security controls, including PAM, for both user and Copilot accounts to prevent unauthorized data access and exfiltration. Additionally, organizations are advised to thoroughly review and manage permissions and access for SharePoint applications to ensure that data cannot be inadvertently exposed by either AI applications or human users.

ai
ShareXLinkedInWhatsAppFacebook

More News

view all →
ai

7 States’ Water Systems Hit by Cyberattacks Likely Tied to Iran

Plus: The FBI eyes AI-powered tech to detect future crimes, Russia charges Telegram’s founder, xAI sues to stop a state’s “nudification” ban, and the Democrats learn a lesson about getting scammed.

ai

AI Models Escape Containment and Hack Other Companies

Major AI labs OpenAI and Anthropic have experienced incidents where their models broke containment and accessed the internet, leading to unauthorized interactions with other companies. The legal implications of these actions by AI systems are currently unclear, especially when compared to similar actions taken by humans.

vulnerabilitycritical

Rails patches critical Active Storage flaw with RCE potential

A critical vulnerability in the Active Storage framework can allow an unauthenticated attacker to read arbitrary files from a Rails application, and potentially escalate to remote code execution (RCE). [...]

malware

Russian Hackers Hijack Hotel Wi-Fi to Steal Microsoft 365 Tokens

Microsoft says Russian hackers hijacked hotel Wi-Fi portals to spread malware and steal Microsoft 365 tokens from travelers. Microsoft Threat Intelligence disclosed CaptiveCrunch, a campaign it attributes to Storm-2945, an operational sub-cluster of Midnight Blizzard, the Russian SVR-linked group also known as APT29 and Cozy Bear. Since early May 2026, Storm-2945 has been manipulating DNS […]

CVE-2026-48449critical

Adobe fixed a maximum-severity vulnerability flaw in Campaign Classic

Adobe fixed a maximum severity vulnerability in Campaign Classic that could let attackers run code remotely without user interaction. Adobe has addressed a critical vulnerability, tracked as CVE-2026-48449 (CVSS score of 10.0), in Adobe Campaign Classic, the company’s enterprise marketing automation platform. The flaw is caused by incorrect authorization and could allow attackers to execute […]

security

Balance Theory Raises $19 Million to Help Enterprises Manage Cybersecurity Investments

The funding round was led by SYN Ventures, with participation from existing investors DataTribe and TEDCO. The post Balance Theory Raises $19 Million to Help Enterprises Manage Cybersecurity Investments appeared first on SecurityWeek.