AGP Picks
View all

Cyfendo targets security risks in AI-built apps for small teams

Oct. 19, 2026
Cyfendo targets security risks in AI-built apps for small teams
By AI, Created 20:01 UTC, Oct 19, 2026, AGP -

Former Google security leader Bryan Vuong has launched Cyfendo to help small businesses validate vulnerabilities and test fixes in AI-built applications. The company is pitching the service during Cybersecurity Awareness Month as more teams move core operations online and face hidden access-control and code risks.

Why it matters: - Small businesses are building more customer portals, order systems and internal tools with AI help. - Those apps can appear to work normally while still exposing private records, invoices or credentials. - Cyfendo is aimed at closing the gap between fast software creation and the security review small teams often lack. - More information is available through Cyfendo's LinkedIn page.

What happened: - Cyfendo used Cybersecurity Awareness Month to highlight security risks in AI-built apps and to promote its application-security platform for smaller teams. - The company said its platform analyzes code, checks whether findings can be exploited, and prepares tested fixes for developers to review. - Bryan Vuong founded Cyfendo after senior engineering roles at Google and Facebook. - Vuong said he founded Cyfendo to help ordinary businesses protect what they have built.

The details: - Google Threat Intelligence Group reported in September 2026 that a suspected attacker used AI to build and run a credential-harvesting campaign from compromised cloud infrastructure in less than six hours. - The campaign used automated scanning and troubleshooting and compromised thousands of third-party credentials. - CISA has named its 2026 Cybersecurity Awareness Month campaign “Securing the Next 250.” - A small business can have a portal that logs in successfully and still leak data if access permissions are wrong. - Cyfendo says developers can connect a GitHub, GitLab or Bitbucket repository, upload project files, or use a command-line workflow. - The platform traces paths from untrusted input to sensitive operations, runs isolated tests on supported findings, prepares a patch and rechecks relevant behavior. - Developers review the evidence and decide whether to accept a change. - Nothing is merged automatically into production. - Cyfendo says it finds an average of four critical-severity issues per codebase in its own scans. - About seven in ten of its findings are rated high or critical. - Cyfendo has also published benchmark results from OWASP test suites. - In September 2026, the company reported 94.06% recall and 96.38% precision on the OWASP Java Benchmark v1.2, which has 2,740 test cases. - On the OWASP Python Benchmark v0.1, which has 1,230 test cases, Cyfendo reported 90.04% recall and 93.78% precision. - The company said its free plan includes 10,000 protected lines of code and five scans per month. - Paid self-service plans start at $99 per month for Starter, then $299 for Growth and $799 for Scale. - Business and Enterprise plans are quoted separately. - Cyfendo also offers application penetration testing for select businesses as a separate service. - The code-security product is publicly available now.

Between the lines: - The pitch is not just about finding vulnerabilities. It is about validating whether a flagged issue is real and whether a fix will break the application. - That matters for small teams, where the same people often build features, answer support requests and handle security work. - Cyfendo's own benchmark numbers are strong, but they are company-reported results and do not prove identical performance in every live system. - The broader message is that AI is lowering the barrier to software creation for defenders and attackers alike. - Verizon's 2026 Data Breach Investigations Report found 7,152 confirmed breaches involving small and medium-sized businesses, with vulnerability exploitation accounting for 26% of initial access methods and third parties involved in 55% of breaches. - Anthropic also published September research describing criminals using AI for reconnaissance, exploitation and handling stolen information.

What's next: - Cyfendo is positioning its free scan as the entry point for teams that want to inspect code before vulnerabilities turn into incidents. - The company expects developers to use the platform to confirm exploitability, prioritize repairs and test that customer-facing features still work after a fix. - As AI-built apps spread, more small businesses are likely to face pressure to add security review to their development workflow.

The bottom line: - AI can help small teams ship software faster, but it can also hide security mistakes that normal testing misses. - Cyfendo is betting that small businesses will pay for a tool that proves whether a vulnerability matters and helps them fix it without handing over control of production code.

Disclaimer: This article was produced by AGP Wire with the assistance of artificial intelligence based on original source content and has been refined to improve clarity, structure, and readability. This content is provided on an “as is” basis. While care has been taken in its preparation, it may contain inaccuracies or omissions, and readers should consult the original source and independently verify key information where appropriate. This content is for informational purposes only and does not constitute legal, financial, investment, or other professional advice.

Sign up for:

SMB & Me

The daily local news briefing you can trust. Every day. Subscribe now.

By signing up, you agree to our Terms & Conditions.

Share this page:

Advanced Search Options

Search for:

Search scope:

Type:

Search in:

Date range:

The last

Sort by:

Sign up for:

SMB & Me

The daily local news briefing you can trust. Every day. Subscribe now.

By signing up, you agree to our Terms & Conditions.