Strategic Cloud Partnerships Lead

The startup helps developers build AI agents that work over iMessage, SMS/RCS, email, and other messaging platforms. It's a bet that consumers will increasingly

Grok matches a cheaper AI on security tests, at nearly 65 times the cost. Artificial Analysis has launched a new Cyber Index for testing AI models on vulnerability discovery, validation and repair but performance comes at a cost. Artificial Analysis has launched a new Cyber Index designed to measure how well AI models find and fix software vulnerabilities, while showing how much each evaluation costs to run. The index combines three cybersecurity evaluations: CWE-Bench-AA, DeepsecBench-AA and CyberGym-E2E-AA. Together, they cover vulnerability discovery, validation and remediation. The launch also establishes the Cyber Index Alliance, with Collinear AI, IBM, NVIDIA and Vercel as founding partners. Collinear AI and Vercel contributed benchmark work, while IBM and NVIDIA provided expert input, according to Artificial Analysis. What counts as a successful fix? Finding a vulnerability is only part of the task. In CWE-Bench-AA, an AI agent receives a real open-source repository and an instruction describing the area of concern, but not the vulnerability's exact location. It must audit the code, identify the weakness and patch it. This comes as a broader move toward giving agents autonomy to execute tasks. A fix counts as successful only when a programmatic verifier confirms that the vulnerability can no longer be triggered and legitimate functionality still works. There is no partial credit or language-model judge deciding whether the patch looks good. Artificial Analysis reports results as pass@1, meaning the percentage of 120 held-out tasks solved on the first attempt. Tasks run in an isolated sandbox without internet access. This creates an important distinction: removing vulnerable code is not enough if the patch breaks legitimate functionality. Similarly, closing one route to a vulnerability while leaving a related weakness open does not receive credit. Artificial Analysis says partial fixes were the main failure mode in CWE-Bench-AA. Excluding refusals and timeouts, 55% of failed attempts fixed the primary issue while leaving a related weakness open. Three tests, not one. The three evaluations measure different defensive tasks. CWE-Bench-AA tests vulnerability identification and remediation across 120 held-out tasks covering the OWASP Top 10 2025 categories and several programming languages. DeepsecBench-AA focuses on vulnerability discovery in open-source application code, with results compared against expert-verified findings. CyberGym-E2E-AA tests whether models can discover, reproduce and patch memory-safety vulnerabilities in C and C++ projects. The launch version contains 131 tasks. Each evaluation contributes equally to the overall Cyber Index score. The index does not ask models to develop working exploits. Performance comes with a cost. The first results show why cost is worth considering alongside the score. Artificial Analysis reports that Grok 4.7 (xhigh) and MiMo-V2.6-Pro both score 56 while the cost per task is $11.67 and $0.18 respectively. GPT-6 Luna (max) scores 53 and the cost per task is $0.12. These are Artificial Analysis' evaluation costs, not estimates of production deployment costs. They are calculated from the tokens used during the evaluations and available model pricing. The comparison is therefore useful for understanding the trade-off between benchmark performance and evaluation cost, but it does not capture infrastructure, human review, monitoring or failure-handling costs. What the Cyber Index shows. Artificial Analysis also reports safety blocks separately. When a model or provider declines a task on safety grounds, that task receives zero credit, although the refusal rate is reported separately. The launch index also has clear limits. It does not cover areas such as incident response, writing new code without introducing vulnerabilities or testing targets without source-code access. For security teams, the useful takeaway is therefore not simply which model has the highest score. The benchmark provides a way to compare defensive capability, task coverage and cost under a defined testing methodology. Artificial Analysis says it plans to expand the index. For now, its results should be read as evidence of how models performed on these specific source-code-based security tasks, rather than as a complete measure of an AI system's ability to handle cybersecurity work. 1 Oct 2026 AMD has agreed to acquire World Labs for $8.2 billion to shape the design of its future hardware, software, and systems. 1 Oct 2026 Instinct raises $1 billion to expand its AI concierge, though users must still review and approve key details before bookings are finalized. 30 Sep 2026 Adobe forecasts AI-referred traffic to US retail sites will grow 130% this holiday season, but says the channel remains modest in overall volume.

Nudge Security Unveils Adaptive Risk Management to Track How SaaS and AI Risk Changes as Usage Evolves After Approval TMCnet News [September 30, 2026] | / | Nudge Security Unveils Adaptive Risk Management to Track How SaaS and AI Risk Changes as Usage Evolves After Approval New capability continuously reassesses risk as employees connect integrations, expand access, and adopt AI tools, closing a gap left by one-time vendor reviews. AUSTIN, Texas, Sept. 30, 2026 /PRNewswire/ - Nudge Security, a leader in SaaS and AI security governance, today announced Adaptive Risk Management, new functionalities that respond to changes in SaaS and AI vendor risk as well as internal usage, helping security teams to more effectively manage risk across a sprawling SaaS and AI attack surface. Most organizations evaluate a vendor once, during procurement, and treat that assessment as settled. But an application's criticality and risk to a specific business isn't static. After the vendor review and approval is complete, employees often connect it to other business apps and AI agents, share more sensitive data, or invite more internal and external collaborators - often without security's knowledge or review. Traditional third-party risk tools are built to monitor a vendor's external security posture and internal security controls applied once. They are not built to reassess or apply additional security controls as enterprise usage changes. Nudge Security does both, continually monitoring vendors' external risk posture and recalculating an application's risk score as internal usage changes. As risk scores change, Nudge Security applies and recommends appropriate mitigating and compensating controls, natively and through third-party security integrations, creating a truly adaptive risk management program. "Nudge provides a relational view across applications, users, integrations, browser extensions, and authentication that surfaces where security actually needs improvement," said Diego Izquierdo, Senior Cybersecurity Engineer, Third Party Risk Management at Mercado Libre. "For example, an organization can have very strong controls in a tool like Slack, but if users connect less-secure third-party apps into it, that can create a breach path without anyone realizing a new risk was introduced. That integrated perspective on risk is wat makes Nudge unique." Third-party involvement in breaches has risen 60% year over year to account for 48% of all breaches, according to the 2026 Verizon Data Breach Investigations Report. Compressed vulnerability cycles in the post-Mythos era stand to accelerate that trend further, and recent supply chain incidents (including those at Vercel, Salesloft Drift, and LastPass) have already shown how quickly access granted to one application can expose an entire environment. Security teams are increasingly asked not just whether a vendor is secure, but what that vendor's access can actually reach inside their environment right now - a question a point-in-time review can't answer on its own. Compounding this challenge, most organizations actively manage only 30-40% of the SaaS and AI tools actually in use, and typically discover two to three times more tools than they expected with Nudge Security. Adaptive Risk Management is built to close that gap by continuously combining what's known about a vendor's security, supply chain risk, and compliance program with what's actually happening inside a customer's own environment. * Automatic app criticality tiering: Every application is automatically classified by the breadth of use across the organization along with the sensitivity of the data it typically handles, using a proprietary AI model that identifies up to 29 distinct data types, so security teams know which applications matter most without reviewing each one manually. * Continuous risk scoring: Each application gets a dynamic risk score built from more than 30 factors, combining a vendor's external security posture with how that specific application is being used inside a customer's environment: approval status, access granted, data touched, MCP connections, who or what is using it, what security controls in place, and what security findings are failing. Scores update automatically as those factors change, rather than waiting for the next scheduled review. That includes signals like whether a vendor's OAuth grants are stale or unused, whether an AI agent or integration is connected to the app, and whether employees are authenticating through single sign-on or with standalone credentials. * A direct line from risk to action: Nudge Security ranks the specific control gaps driving an app's risk score - such as enabling SSO or closing a stale OAuth grant - by how much closing each one would reduce that risk, and in many cases security teams can act on it directly from the risk panel. The platform also adapts its own monitoring as an app's importance changes: mark an app business-critical, and Nudge automatically begins surfacing additional findings for it - like password reuse or weak authentication - that weren't being tracked before. Risk scores update automatically as those factors change - and Nudge Security customers using strong compensating controls like SSO and MFA can reduce an app's residual risk by as much as 60%. "Too many organizations still treat third-party risk as an annual exercise, while in reality, the conditions that impact risk change every single day," said Jaime Blasco, co-founder and CTO of Nudge Security. "An application approved for a handful of employees becomes entrenched across the business. Someone connects an AI tool to sensitive data or grants an integration broader access. A breach at one of your vendors suddenly turns that access into a potential path into your environment. Your last vendor assessment won't tell you what's at risk now. Fog Computing World built Adaptive Risk Management to connect changes in usage, access, and a vendor's security posture so security teams can continuously reassess their exposure and act as it changes. Nudge Security built its risk model on its own data rather than third-party scoring services, drawing on a self-populating database of more than 250,000 SaaS and AI vendor security profiles. That means Adaptive Risk Management can assess an application from the moment it enters an environment, without requiring vendor cooperation, manual data entry, or even prior knowledge that the application was in use." Availability Adaptive Risk Management is available now to all Nudge Security customers. About Nudge Security Nudge Security delivers SaaS and AI security governance at the Workforce Edge, where employees make thousands of technology decisions daily. Its automated, policy-driven guardrails reach employees when and where they work, enabling rapid technology adoption while minimizing risk and sprawl. Through unrivaled discovery capabilities, AI-driven risk insights, and behavioral science-based engagement, Fog Computing World make security a natural part of how modern work gets done rather than an obstacle to innovation. Nudge Security was founded in 2021 by Russell Spitler and Jaime Blasco and is backed by Cerberus Ventures, Ballistic Ventures, Forgepoint Capital, and Squadra Ventures. Media Contact: Danielle Ostrovsky Hi-Touch PR [email protected] View original content to download multimedia:https://www.prnewswire.com/news-releases/nudge-security-unveils-adaptive-risk-management-to-track-how-saas-and-ai-risk-changes-as-usage-evolves-after-approval-302893378.html SOURCE Nudge Security [ Back To TMCnet.com's Homepage] |