As the security industry is reshaped around AI, M&A activity is intensifying.

Data platform company Databricks is acquiring Panther Labs, a startup that provides an AI-based cyberattack detection platform. Panther is the third cybersecurity startup Databricks has acquired this year. Databricks previously acquired data encryption specialist Antimatter and incident response automation platform SiftD. Identity security company SailPoint Technologies is acquiring AI agent security startup Entro Security.

• Databricks carries out another security M&A, this time buying attack detection platform Panther Labs • SailPoint acquires AI agent security startup Entro

Global big cloud companies are also moving faster on security.

Amazon Web Services has launched its security platform AWS Continuum, which uses AI to find software vulnerabilities, validate them and recommend fixes. Google Cloud is supporting local data residency for Google Security Operations, or Google SecOps, in the Google Cloud Seoul Region, which it refers to as the Seoul region.

• AWS unveils AI security platform Continuum, automating code vulnerability detection, validation and fix recommendations • Google Cloud launches Google Security Operations platform in its Seoul region, supporting data sovereignty for Korean companies

It also compiled moves and issues around security by companies at home and abroad.

SentinelOne made its Purple AI autonomous threat investigation feature, Purple AI Agentic Investigation, available to all customers. It automatically activates when threats exceeding customer-defined criteria are detected, even if an analyst does not start an investigation. Purple AI identifies and assesses threats, then carries out responses based on the results. Okta is expanding its partnership with Google Cloud with a focus on strengthening AI agent and browser security. SoftBank Group launched a new cyber security product based on an OpenAI model, Patching as a Service. Beyond Identity launched its AI agent security platform Ceros. Beyond Identity provides an identity and access management, or IAM, platform that allows companies to manage employees' application access permissions.

• SentinelOne applies autonomous threat investigation feature to Purple AI • Okta and Google Cloud expand partnership, strengthening AI agent and browser security • SoftBank launches new cybersecurity product based on OpenAI model • Beyond Identity: "Ceros controls access for AI agents"

Hanssak launched SecureGate V4.0, a data transfer (network linkage) solution between networks, as part of efforts to target the market for transitions to the National Network Security Framework, or N2SF. Piolink completed patent registrations in South Korea, the United States and Japan for a technology that implements micro-segmentation using its TFront security switch. The company said it offers a realistic and cost-effective alternative for institutions preparing for an N2SF transition. The Korea Internet & Security Agency, or KISA, selected 6 consortiums made up of the Ministry of Science and ICT, national and public institutions and security companies. It is pushing ahead in earnest with its 2026 project to support adoption of the national network security framework. WINS TechNet signed a memorandum of understanding with Chung-Ang University’s AI Research Institute on research into AI-based security technologies. Application security company Sparrow plans to officially launch Coding Sparrow MCP soon.

• Hanssak launches new network linkage solution SecureGate V4.0, responding to N2SF transition • Piolink targets N2SF market with switch-based micro-segmentation technology • KISA steps up support for public institutions to adopt N2SF • WINS TechNet to cooperate with Chung-Ang University AI Research Institute on joint AI security research • Sparrow to introduce coding AI security product Sparrow MCP

Tens of thousands of Fortinet firewalls and VPNs at large security firms were attacked. The attack did not exploit a newly discovered unknown vulnerability. It stemmed from companies not changing firewall passwords or not checking whether credentials for sensitive systems exposed to the internet had already been leaked.

• Large-scale breach of Fortinet firewalls and VPNs, exploiting leaked passwords

Cybersecurity chiefs and researchers at U.S. tech companies urged the Trump administration to roll back regulations on Anthropic’s latest AI model. They argue the measures are dealing a bigger blow to defenders than to attackers.

• About 40 security experts: "Regulation of Anthropic model hits defenders more than attackers"