< ciso
brief />
Tag Banner

All news with #ai security tag

1155 articles

Signal Adds Protections to Block Siri and Gemini AI

๐Ÿ”’ Signal has introduced new protections that stop AI assistants such as Siri and Gemini from capturing private conversations on iOS 27 and supported Android devices. The safeguards also block screenshots on iOS. Signal president Meredith Whittaker warns that OS-level AI agents remain an existential threat to the messenger, and she urges Apple, Google, and Microsoft to document every AI access pathway and let developers opt out.
read more โ†’

Bell Canada Confirms Agentic Attacker Stole Customer Data

๐Ÿšจ Bell Canada has confirmed that an agentic threat actor breached its systems on November 17 and exfiltrated enterprise customer data. The intruder accessed a pre-production Maximo environment used to order and provision telecommunications circuits. Exposed data includes business contact details, ticket titles, and some IP addresses, but not credentials. Bell says it contained the incident and that customer services were not affected.
read more โ†’

European Commission Confirms AI-Assisted Cyberattack

๐Ÿ›ก๏ธ The European Commission has confirmed that it was hit by an AI-assisted cyberattack in late October that disrupted several of its services. Applications were temporarily taken offline as a precaution, and most have since been restored. Officials say there is no evidence of a data breach affecting staff, though the investigation is still ongoing. The attacker has not been identified.
read more โ†’

Critical Cisco Expressway Flaw Exploited in Russian Breach

๐Ÿšจ Attackers exploited CVE-2026-76466, a critical (CVSS 9.8) flaw in Cisco Expressway, to breach a Russian organization soon after a proof-of-concept exploit surfaced. The bug lets unauthenticated remote attackers run commands as root. F6 reports the intruders moved laterally using custom Python scripts that may have been built with AI assistance. Cisco has released patches, and defenders are urged to update and audit their environments.
read more โ†’

European Commission Confirms AI-Generated Cyberattack

๐Ÿ›ก๏ธ The European Commission has confirmed that a cyber campaign misusing a publicly available AI tool forced it to take several applications offline. The incident was detected on 22 October and confirmed on 26 October. Most affected systems have since been restored, and no evidence of a staff data breach has been found so far. CERT-EU reportedly linked the activity to the AI-powered tool ARTEX.
read more โ†’

AI Math Advances Spark Debate Over Public-Key Crypto Risk

๐Ÿ” Johns Hopkins cryptographer Matthew Green has warned that AI's growing strength in solving mathematical problems could shake confidence in today's public-key encryption. Other experts, including Markku-Juhani Saarinen and Alan Woodward, take the concern seriously but urge calm, pointing to the ongoing PQC transition and cryptographic agility as the sensible response rather than any hasty rip-and-replace.
read more โ†’

European Commission Confirms AI-Driven Cyberattack

๐Ÿ›ก๏ธ The European Commission has confirmed a cyberattack first detected on October 22, describing it as an AI-generated cyber campaign. Reports indicate the attackers used ARTEX, an open-source AI hacking platform, and disrupted personnel systems such as payroll and medical reimbursements for more than 10 days. Investigators have found no evidence of a broad data breach, and the probe remains open.
read more โ†’

European Commission Confirms AI-Driven Cyber-Attack

๐Ÿšจ The European Commission has confirmed that an AI-driven cyber-attack forced several of its applications offline. The intrusion was detected on October 22 and confirmed on October 26, prompting mitigation and an investigation. Officials say no evidence of a data breach affecting staff has been found so far, and the EDPS has been notified. Reports link the activity to the Chinese AI tool ARTEX.
read more โ†’

Cisco Expressway Flaw Exploited Soon After PoC Release

๐Ÿšจ Attackers exploited the critical Cisco Expressway flaw CVE-2026-76466 (CVSS 9.8) to breach a Russian organization shortly after a public PoC appeared, according to F6. The bug allows unauthenticated remote attackers to run commands as root on clustered Expressway Core and Edge devices. Cisco fixed the issue in Expressway X15.5.20, and no workarounds exist. F6 also found signs that an AI assistant helped the attackers tweak their tools during the intrusion.
read more โ†’

Closing the Governance Gap in Open-Weight AI Deployments

๐Ÿ›ก๏ธ Open-weight AI is going mainstream, with an estimated 42% of surveyed enterprises running such models in production. Self-hosting offers greater control over data and less provider dependence, but it also shifts security responsibilities onto the organization. Risks include trojanized model files, unclear model provenance, manual updates, and a larger attack surface. The article recommends clear ownership, verified sources, rigorous testing, and isolation before deployment.
read more โ†’

Palo Alto Networks Buys Applied Compute for $4.5B in Stock

๐Ÿค– Palo Alto Networks has agreed to acquire Applied Compute in an all-stock deal valued at $4.5 billion. The move will let the company train, deploy and secure open-weight AI models for enterprises in the public cloud or sovereign environments. CEO Nikesh Arora argues that securing AI agents requires helping build the supervisory models behind them. Analysts, however, urge CIOs to demand independent testing and clear exit terms.
read more โ†’

Senators Urge White House to Build Defensive AI Cyber Tools

๐Ÿ›ก๏ธ Sens. Maria Cantwell and Mike Rounds have urged the White House to launch a national effort against AI-enabled cyber threats. In a letter to DNI Jay Clayton, they warned that defenses are not keeping pace with AI's ability to find and exploit software flaws. They want faster deployment of defensive AI tools and security standards for autonomous AI agents, with CISA leading adoption across critical sectors.
read more โ†’

Vero Security Raises $11M for AI Security Engineering

๐Ÿ’ฐ Vero Security, founded in 2025 by IDF 8200 alumni, has secured $11 million in seed funding led by TLV Partners. The company is building a platform for AI Security Engineering, a new discipline focused on securing the AI tools and agents that employees increasingly adopt in-house. Proceeds will go toward product development and expanding its go-to-market efforts in the United States.
read more โ†’

Survey: AI Pentesting Adoption Outpaces Testing Cadence

๐Ÿ›ก๏ธ A new Intruder survey of 500 security and engineering leaders in the US and UK finds that AI pentesting is already used across the environment by 55% of organizations. Code ships daily at 68% of firms, yet 47% go over a month between full pentests, and only 4% feel their coverage is sufficient. Data privacy and production safety, not output quality, are the leading adoption concerns. Some 90% agree the annual pentest is dead.
read more โ†’

Securing AI-Generated Code in Enterprise Software

๐Ÿ”’ Organizations face rising risks as AI-generated code proliferates across teams. AI coding assistants accelerate development but often produce insecure outputs, so CISOs must treat AI-written code as untrusted input and enforce automated scans, editor checks, and human review. Practical steps include approved tool lists, specialist routing for high-risk changes, documented ownership of AI-assisted commits, and training for non-traditional developers.
read more โ†’

CrowdStrike Adds Expert-Led MDR to Falcon Guardian for AI

๐Ÿ›ก๏ธ CrowdStrike has made Falcon Complete for Falcon Guardian generally available, pairing AI runtime visibility with 24/7 managed detection and response. The service helps analysts tell malicious agent behavior apart from legitimate workflows by tracing activity from prompt to endpoint impact. It combines adversary intelligence, threat hunting, and customer context to guide remediation across endpoints, identities, cloud, and SaaS.
read more โ†’

Google and Anthropic accelerate vulnerability disclosure timelines

๐Ÿ”’ Google Project Zero is cutting its vulnerability disclosure deadline from 90 to 60 days in February and to 30 days in January 2028, and Anthropic is moving to 30 days for high/critical issues and 45 days for others starting in January. The change responds to AI-driven advances that radically shorten the time attackers need to find and weaponize flaws, including automated pipelines that reverse-engineer patches. Project Zero is also removing its 30-day "soak time," abandoning severity-based thresholds, and will no longer wait for vendor patch verification before publishing details.
read more โ†’

Anthropic Shortens Vulnerability Disclosure Windows

๐Ÿ›ก๏ธ Anthropic is cutting its coordinated vulnerability disclosure window from 90 days to 30 days for high- and critical-severity issues, and to 45 days for all others. The change reflects how quickly AI is improving at finding and even fixing bugs, which makes independent rediscovery more likely. The new timelines apply to reports made on or after January 1, 2027.
read more โ†’

Adversaries Scale Attacks with Agentic AI Frameworks

๐Ÿ” Google Threat Intelligence Group (GTIG) and Mandiant report that adversaries are moving from basic prompting to autonomous multi-agent frameworks, automated exploit pipelines, and LLM-focused monetization such as LLMJacking. These developments compress attack timelines, scale intrusions to machine speed, and introduce new supply chain and IO risks. Google outlines defensive measures, model safeguards, and operational mitigations to detect and disrupt agentic threat activity.
read more โ†’

AI-driven research accelerates vulnerability discovery

๐Ÿ” A new Forescout report, Finding Critical Infrastructure Bugs with Off-the-Shelf AI, shows how publicly available AI models accelerated vulnerability discovery in 10 popular TCP/IP stacks. The team used an AI-assisted harness to cut analysis from roughly nine months to 90 hours, uncovering 99 vulnerabilities including 66 zero days. The study highlights cost efficiencies and cautions that AI aids but does not replace expert researchers, and recommends risk prioritization, device isolation, faster patching and stronger procurement requirements.
read more โ†’