Black Hat 2025 is at the stretch of its home, collecting thousands of security professionals to discuss the latest development in adverse tradecrafts and cyber security defense.
Security leaders and teams discovered AI-operated dangers and innovations, focusing on the autonomous AI agents and Chhaya AI, acceleration of identified attacks and significant importance of hardware-tier security and supply-series security.
conference In addition, new products and updated announcements include a dizzy number, which is mostly focused on (surprise) AI. Here is a look at some of them.
Sukardar launched agent Threat Intelligence
New agents of Sukardardar in their own Agentic danger intelligence The platform understands the context of the danger, identifies the appropriate courses of action, and autonomally trigger the reactions. Agents are experts of different hazards, such as fishing, IP exposure, or credential leaks, and can be mixed and matched and optimized.
Sukardar is also starting what she first calls cyber security AI marketplace, where security teams can browse, buy and manage agents.
Snyk secures AI from installation
Snyk’s new platform capacity, Initially safeThe real -time security scanning involves which begins at the moment of code generation or execution. It provides visibility in generic AI, agentic and model reference protocol (MCP) components in software, and also offers a new, experimental scanner to detect AI-specific MCP weaknesses.
Safe AI installation is now available in Early Access.
AIRMDR AI SoC Tier -1 handles the majority of alert triaies
AIRMDR says its new AI SOC Platform Tier -1 alert automates more than 90% of the tries. Security teams can analyze the basic causes of one-click, sub-5-minute analysis and autonomous reactions. Designed for managed security service providers (MSSPS), its multi-friendly operation customers keep the data isolated by trying and checking at the center. The platform provides 200 -plus indigenous integration and full audit trains.
AIRMDR also introduced a “Free Forever” scheme that supports three data sources and 100 alerts per week.
Deskope manages AI agents
Deskope Identification Management for Agent Identity Control Plain Institutes Policy-based governance, auditing and AI agents and MCP environment.
The platform gives security teams the ability to restrict agents access within specific third-party equipment and allows them to apply policies based on user roles. Monitoring and auditing capabilities help users to spot errors and misconceptions and identify “evil agents”. The AI life cycle management agent provides behavior and visibility in the link with human users.
Cyra secures all types of AI
Cylocardial AI mentor Protects any type of AI, such as from public equipment such as Chatgpt to embedded software-e-service (mother-in-law) model and proprietary platforms. The platform has two main products: AI-SPM (Safety currency management), which provides inventory of all AI assets; And AI runtime protection, which monitors and reacts in real time. The AI Guardian also has Omni AI, a new condensed AI device that can analyze millions of enterprise records in seconds, and creates a safety reports to help relieve dangers.
AI-SPM and Omni AI are available in private bets; AI Runtime Protection is available for early access.
Netscope a copilot
Netscope forest Now the zero trust network access (ZTNA) and a Netscope provides a copylot for the MCP server. AI-Operated Copilot optimizes ZTNA, automatically recommends policies for newly-fed applications, and configures apps and policies. The equipment is now usually available.
NETSKOPE MCP allows popular LLM like cloud, microsoft copilot, Amazon Bedrock and Google Vertex to interact with NETSKOPE management API to get reference to improve the server workflows. The server includes several examples of landscapes, including event analysis and position and insider -risk analysis.
AI-Invented Summary of Flashpoint for search and investigation
Flashpoint ignited It now includes the AI summary of search and investigation. The search summary is directly included in the workflow; With one click, teams can gain insight into discussion with dark web forums, social networks and chat platforms. These are provided in footnoted snapshots.
The AI probe summary is created in ignition, automatically produces reference-supported summary. Users can reproduce, abbreviations and share probe conclusions through PDF or platext files. These conclusions remain updated as an investigation.
Cyware also includes mcp
Cyware MCP Server Integrates lLM directly into the workflows to detect real -time reference and control and provide control in the response. The MCP server is included in the Quarterback AI of the Cyware, which features automated summary and smart passing, where embedded AI agents surface individuals of compromise (IOCs), Adversary TTPS, Malware and Visnebilations. It also provides recommendations from danger intelligence and alert.
Palo Alto Network expands Cortex Cloud
Palo Alto Network Cortex cloud It is now an application security asana management (ASPM) and an open appsec partner ecosystem that shares proved equipment and integrates findings from industry-agronic scanners. Cortex clouds automate significant risks and weaknesses in cloud surfaces, and automatically map the weaknesses and root them for the respective developers. The findings are correlated to the surface of the danger so that the SOC has the visibility of detecting, prioritizing and responding to the SOC.
Crowdastric Falcon Shield expands support for hundreds of apps
Falcon shield Now OpenAI Cuts Enterprise Compliance is integrated with API to help teams to view and operate GPT-based and Codex agents. This expands support for 175-Plus Saas apps.
Falcon Shield maps agents to detect access to its human creators, control privileges and find safe identity; Detects risky behavior; And automatically there are danger through Falcon fusion, no-codes safety orchestration, automation and response (SOAR engine).
Appomni gives app control to teams
Appomni has released Three new product package To control enterprises on their mother -in -law and AI apps. A “Foundation” starter package shadow detects shadow and AI, app users and permissions, and dangers and suspicious activities. An advanced tier provides a more sophisticated danger detection and mother -in -law control, while an enterprise package provides “the highest level of safety maturity”.
The platform also includes the danger detection and support for 30 new AIs and mother -in -law applications, including the Chatgpt Enterprise, Cloud, Cisco umbrella and safe access, and gong.
Tenable AI allows exposure teams to manage and control risks
Created in tenable forest, new Tenal AI Exposure The enterprises help the enterprises to see, manage and control the risks presented by enterprise tools such as Chatgpt enterprise and enterprise use of microsoft copilot. It identifies all AI uses (whether or not), preference risks, and safety railings and organization policies.
Menlo secures storage and controls browser
New Meno safe storage Provides complete access to users to files. However, they are not allowed to save their own devices locally; All file interaction clouds remain limited within the cloud environment. This allows contractors and other third parties to transfer files between sensitive documents and employees safe to reach safely, without exposing them to the endpoint.
The new adaptive web of Meno provides cloud-based browser control. Distributed through Menl’s cloud service, allows platform teams to redress the capabilities of tailor users, redesign, blocks and redirect pages, safe search and disable the password field. In browser sessions, modules can be adapted and deployed depending on individual users or groups.
