Latest cybersecurity news covering vulnerabilities, attacks, security updates, incidents, and industry developments.
3799 articles
The Senate Judiciary Committee advanced the NO FAKES Act, which would give people rights over AI-generated replicas of their likeness and voice. Supporters cite harmful deepfakes and scams; critics warn of overbroad takedowns of lawful speech.
As part of Operation Endgame, international law enforcement cleaned SocGholish from 14,971 WordPress sites and took 106 related servers and domains offline. The malware uses fake browser updates to deliver additional payloads.
The roundup covers attacks abusing browser extensions, AI chat features, developer tools, and OAuth flows, alongside a Cisco SD-WAN zero-day update, npm supply-chain threats, and new security policies.
Accenture plans to take a majority stake in Dragos for $3.25 billion and acquire runZero and NetRise, expanding into OT security software. The deals await regulatory approval and are expected to close in August or September.
Researchers report that TeamPCP compromised more than 1,000 open-source packages in under four months, abusing CI/CD workflows to steal credentials and spread malware. Mini Shai-Hulud’s source code was also released for others to reuse.
Censys found about 8,500 internet-exposed REDCap instances, with only 1.18% running the latest version as of June 16, 2026. The finding is significant given reported targeting of legacy servers by China-linked actor UNC6508.
Accenture announced a majority investment in Dragos and acquisitions of runZero and NetRise, valuing the combined enterprise at about $4.175 billion. The companies aim to combine OT threat detection, asset discovery, firmware analysis and supply-chain visibility.
The article outlines how U.S. executive actions and CISA’s BOD 26-04 shift federal cybersecurity toward risk-based remediation, AI-enabled defense, stronger infrastructure protections, and updated requirements for contractors and critical infrastructure.
Microsoft says it fixed an issue that caused June 2026 security update KB5094122 to fail on some Windows Server 2016 systems lacking the previous KB5087537 update, with 0x80070002 or FILE_NOT_FOUND errors.
India told the Delhi High Court Telegram was warned about alleged NEET-UG 2026 paper-leak channels and said it could not proactively detect them. The ban remains in force pending a ruling; reports of a BGP route leak and Jio's role are disputed.
Operation Endgame took down 106 servers and domains and remediated 14,971 compromised WordPress sites tied to SocGholish. Infoblox telemetry found widespread exposure, but few customer networks showed signs of reaching the malware’s later stages.
Israeli cybersecurity startup Dream raised $260 million at a $3 billion valuation. It says the funds will accelerate deployment of its Sphere, Hero, and Atlas platforms for national cyber defense, vulnerability hunting, and sovereign AI.
Sen. Mark Warner says CISA staffing cuts and the loss of federal support for MS-ISAC are weakening cyber assistance to state and local operators. He introduced legislation to fund MS-ISAC and requested CISA staffing and service-delivery records.
Researchers report that a Russian-speaking group targeted 73,932 Fortinet firewall URLs across 194 countries, using exposed configurations and offline credential cracking. They say some intrusions reached internal networks, including a Turkish defense contractor.
NCSC chief Richard Horne said the agency handled over 200 incidents affecting UK critical infrastructure in the year to May, about 75% believed linked to state actors. He warned of adversary footholds and potential AI-assisted exploitation by 2028.
Check Point Research reports an unknown actor used fake reviews, inflated downloads and coordinated platform activity to promote malicious tools hiding a Rust clipboard hijacker. It replaces matching crypto wallet addresses on Windows and macOS with attacker-controlled,
The NCSC says a global campaign targeted internet-facing FortiGate and VPN portals using brute-force, dictionary and credential-stuffing attempts, and urges organisations to investigate possible impact and follow mitigation advice.
Google plans to use IP addresses for ad measurement and personalization in the EEA, UK and Switzerland from on or shortly after Aug. 3, 2026, raising consent and fingerprinting concerns under privacy rules.
A temporary Indian Telegram ban coincided with an AS18101 route hijack that leaked globally through FLAG Telecom, disrupting access in the UAE. Analysts confirmed the route leak but said evidence does not establish deliberate sabotage; MTProto proxies can restore access
This weekly roundup covers the patched SearchLeak flaw in Microsoft 365 Copilot, calls to restore access to Anthropic models for defenders, security-sector funding and acquisitions, and alleged police misuse of Flock license plate data.
1Password announced its acquisition of Apono, whose just-in-time access governance covers human, machine, and AI-agent permissions. Calcalist reported a $250 million–$300 million deal value; neither company officially disclosed financial terms.
Tenet Security raised $6 million in seed funding for a platform it says monitors AI agents at runtime and can stop harmful actions. The company also reported findings from its research and claimed detections in early deployments.
Google and Mozilla released browser updates fixing more than 70 vulnerabilities, including critical and high-severity memory-safety flaws that could potentially enable code execution. Google said it had no information that the Chrome flaws were exploited in the wild.
Defused reports in-the-wild targeting of three patched FortiSandbox flaws. Separately, SOCRadar says the FortiBleed campaign compromised more than 30,000 Fortinet firewalls to collect credentials; the actor is unattributed.
