From AI strategy to secure production.
Security experts with 20+ years in cyber and a decade building and securing AI. One team takes you from idea to engineering, integration and secure deployment.
One team, from idea to secure deployment.
Most AI vendors either build or secure. We are security experts who also engineer, so protection is designed in from the first workshop instead of bolted on before launch.
- Senior security experts on every engagement
- Strategy, engineering, integration and security under one roof
- Accountable from first idea to production and beyond
AI is our focus. Information security is our foundation.
Information security servicesHow to start
Three fixed-scope ways to begin. Each ends with a clear report and a plan for what comes next.
AI security assessment
Typical duration: 2–3 weeks
One AI assistant, agent or workflow reviewed end to end: threat model, adversarial tests and data-exposure checks.
Ask about thisYou receive
Findings with evidence, a prioritized fix plan and a board summary.
Microsoft 365 Copilot readiness
Typical duration: 3 weeks
Find what Copilot could expose before your employees stumble on it, then fix permissions, labels and DLP.
Ask about thisYou receive
An oversharing map, a label and DLP plan, and rollout guardrails.
Penetration test
Typical duration: 1–4 weeks, scoped
AI applications, web and API, network and cloud, or physical intrusion and social engineering, scoped to what matters most.
Ask about thisYou receive
A report with evidence and clear fixes, and a retest once you have fixed them.
Not sure where to start? Begin with a 30-minute discovery call.
Security that follows the work, not only the data.
Access to data doesn't authorize every AI action. We secure the whole workflow, then prove the controls hold with AI penetration testing.
What you receive
Every assessment ends with findings your engineers can fix and a summary your board can read.
Every engagement includes
- Documented findings with clear fixes
- A one-page board summary
- A retest of fixed findings
Support bot can be talked into issuing refunds
- Mapped to
- OWASP LLM06 Excessive Agency, MITRE ATLAS AML.T0051 LLM Prompt Injection
- Evidence
- A crafted customer message made the agent call the refund tool without approval.
- Fix
- Require human approval for refunds and limit the tool to read-only order lookups.
- Retest
- Passed after the fix
Standards we map our work to
- OWASP Top 10 for LLM Applications
- MITRE ATLAS
- NIST AI RMF
- ISO/IEC 42001
- ISO/IEC 27001
- EU AI Act
- Israel Privacy Protection Law, Amendment 13
AI use cases for the enterprise
Six areas where AI pays off for enterprises. Pick one to see examples and how we deliver it safely.
Turn a business idea into a secure AI application your teams actually use, from first prototype to production.
AI handles the routine steps of HR, procurement, IT and finance processes, while people approve what matters.
Employees ask in plain language and get answers from your own policies, procedures and documents.
AI reads documents, extracts the key data and cross-checks it, sending only the exceptions to your experts.
Service and sales teams get AI that summarizes, drafts and recommends, and customer-facing assistants launch safely.
Know where AI is used across the company, set clear rules, and prove your AI is safe before and after launch.
Who you'll work with
Senior people on every engagement, from strategy to the test report.
Tony Voloshko
Founder, AI and cyber security
Tony founded 3Lens after more than 20 years securing complex systems in banking, online gaming and other regulated industries, including as CISO of a startup building an AI product. He works hands-on in every engagement, from the first architecture workshop to the test report your board reads.
And the rest of the team you will work with
3Lens is led by Tony, and these are the senior specialists who make up the company. Depending on your project you will work directly with them, so you get the expertise the problem actually needs.
Former CISOs and security leaders
People who have carried responsibility for security in banks, gaming platforms and critical infrastructure.
AI engineers and ML practitioners
Builders who have shipped LLM assistants, RAG systems and agents into production, and stayed to operate them.
Offensive security researchers
Red teamers and penetration testers who break applications, AI agents, networks and people for a living.
Privacy and AI regulation lawyers
Counsel who work daily with the EU AI Act, the GDPR and Israeli privacy law.
Cloud and Microsoft 365 specialists
Architects and identity experts who secure Azure, Entra, Purview and Copilot at scale.
Data and detection engineers
People who build the logging, monitoring and response that keeps an AI system observable after launch.
Built for regulated organizations
Confidential by default
NDA before scoping. Findings are encrypted, shared on a need-to-know basis and deleted on request.
Independent testing
For assurance work, testers are separate from the people who built the system.
Local and international rules
Work mapped to Israeli privacy law, sector regulators' requirements, and EU and international AI standards.
Hebrew and English
Workshops, reports and board summaries in the language your teams work in.
Questions we hear often
How does an engagement start?
With a 30-minute discovery call. We agree on one workflow or system to look at first, then send a fixed scope and timeline.
Do you only advise, or also build and test?
All three. We design and build AI solutions, integrate them with your systems and test them the way an attacker would, so security is part of the work from day one.
Do you also handle information security that isn't about AI?
Yes. AI is our focus, but information security is where we come from: risk assessments and hardening, penetration testing of AI and classic systems alike, compliance programs such as ISO 27001 and SOC 2, cloud security and data protection.
Can the team that builds our AI also test it?
For assurance and penetration tests we assign testers who were not involved in building the system, so the results are independent.
Do you run physical penetration tests and social engineering?
Yes. Alongside AI, web, network and cloud testing, we run physical intrusion tests and phishing campaigns, always with written authorization and agreed rules of engagement.
Which frameworks do you work with?
We map findings and controls to the OWASP Top 10 for LLM Applications, MITRE ATLAS, NIST AI RMF and ISO/IEC 42001, and to the rules that apply to you, such as Israel's Amendment 13 and the EU AI Act.
Do you work with companies outside Israel?
Yes. We work in Hebrew and English, with Israeli enterprises and with international companies.
Bring us one AI challenge that matters.
Most engagements start with a discovery call or a focused review of one workflow. Tell us what you are working on.
Or reach us directly
Derech Menachem Begin, Tel Aviv-Yafo, IsraelWorking with clients in Israel and worldwide.



