Without any reservations everyone that I know in deep tech or the ivory towers of research institutions are speaking with one voice: AI will change how we look at every opportunity and every challenge.

This note will be sent to our customers monthly to advise as to what SilverSky is doing to protect our customers as well as drive your compliance objectives.

AI? What are we doing about it?

We have engaged with AI on a variety of fronts and also have been providing AI enabled Analytics for many years, our knowledge of these systems is up to date. Our focus has always been on providing transparent explainable AI that allows customers and compliance monitors to see clearly how conclusions are reached.  However, things are moving fast and we have been accelerating our various uses of AI to provide better services as well as understanding the threat of AI  to your systems and their operations.  Regrettably, AI is accelerating every aspect of cyber-attacks, but the potential for AI enhance cybersecurity is more significant.  Powerful AI cyber defender advantages include: operating at scale, collaboration, and sharing information.

With the advent of the New Anthropic Cyber Security (pre-released 4/7/26) model we are tracking the impact of this system being used for good purposes as well as bad.

The Good

Software built by either machine or human contain conditions or errors which can be used to exploit the systems where software is used. With the release of the Anthropic Cyber model, the AI has learned what to look for to tighten security by identifying conditions that have a probability of exploit. This will allow software developers as well as software corporations (Microsoft) to better secure their systems.

The Bad

As everyone immediately understands, bad guys will have access to these systems as well – exploit probability will increase as well as activity.

What SilverSky is doing?

1. SilverSky will be releasing this summer a Compliance product (which will be part of our Lightning Offering) bases on the MITRE ATLAS (Adversarial Threat Landscape for Artificial-Intelligence Systems) where we will be looking at the telemetry from your systems within this attack framework in addition to the other compliance driven frameworks we are currently monitoring (Financial, HeathCare, and others).
2. Double checking all requests from customers to assure we are not being socially engineered. We have always had a method of checking the veracity of change requests or information requests and we are strengthening these systems against the advancements within AI
3. Added a world renowned AI expert, Dr. Jeremy Kepner, to our staff who is assisting in our mathematics and management of the AI Threat horizon.

What Can you Do?

1. Double Down on two factor authentication systems with all customer requests – this is critical
2. Within your vendor reviews ask each of your vendors as to their process in developing their software and their position on AI enabled threats.
3. Train employees to recognize new AI-driven threats
   • Deepfakes
   • Hyper-personalized phishing
   • Voice/video impersonation
   • Foster a skeptical-first / security first culture prioritizing validation over basic or blind trust
4. Strengthen your data security posture
   • A lot of AI attacks rely on harvesting and exploiting organizational data, so having a renewed focus on least-privilege access, data encryption, and expanded monitoring for data exfiltration is critical.
5. Review MFA methods and move to Phishing Resistant MFA options

Schedule a call with SilverSky to address concerns, put together an AI mitigation frame work and proactively address this evolving threat on our data security horizon.