PayPal buys into bio-inspired security with CyActive

happygeek 2 404 Views

I don't usually write about acquisitions and all that financial stuff, but news that PayPal has acquired CyActive caught my eye as apparently this brings the promise of 'bio-inspired predictive security' into the online payments provider threat protection mix. Which made me think, just what the heck is bio-inspired predictive security when it's at home, and why has PayPal bought into it?

cyactive.jpg

My first port of call in trying to get a line on this was the on the thing. "While we have industry-leading fraud models and verification techniques, and a world-class security team" James Barrese, Chief Technology Officer and Senior Vice President, Payment Services, PayPal says "we’re always looking for ways to make our systems even more secure." Which is where the CyActive acquisition comes in, along with the establishment of a security center in Israel that will "tap into the country’s cutting-edge technology and top cybersecurity talent." CyActive being part of that tapping into process, being an outfit which specializes in predictive technology that focuses on how malware will develop and by so doing adds an element of future-proofing (or at least that's the idea) to PayPal security measures.

OK, so what does CyActive actually do then? Good question, and according to the company itself the answer is "forecasts how hackers will evolve today’s malware into tomorrow’s advanced threats, by applying bio-inspired algorithms and a deep understanding of hackers’ behavior, considerations and constraints." Which is about as clear as mud when it comes to explaining what these bio-inspired algorithms actually are. I mean, I get the predictive stuff which looks at malware behaviors and what the cyber-criminals behind it expect to achieve. I'm less clear on how this is turned into accurate predictions of the variants that will emerge from any given malware sample. I guess that by bio-inspired what is actually meant is 'evolutionary best guess based on past experience' to be honest. That doesn't sound quite as impressive, of course.

Not that I'm knocking what CyActive claims to do, the company was founded by a couple of clever bods who apparently were part of the Israeli Defense Force (IDF) which, or at least the 8200 intelligence unit bit of it, appears to have quite a long history of spinning off hi-tech and highly clever start-ups. If, as CyActive claim, it can (accurately) predict hundreds of thousands of future malware derivatives in a matter of hours and then use these to program detectors that proactively anticipate and prevent attacks on network and endpoint devices, that has to be a good thing for PayPal users. I'm certainly all in favor of taking a proactive rather than purely reactive approach to security, and applaud PayPal for seeing the value in that.

Dani AI

Generated

To clarify ’s question: “bio‑inspired predictive security” is a practical label for techniques that borrow ideas from natural evolution to anticipate how malware families might change. The core idea is not magic — it’s automated generation and evaluation of plausible variants from known samples, plus extraction of behavioral signals that generalize beyond a single signature.

A typical pipeline looks like this: start from known malicious seeds, apply controlled mutation/obfuscation operators to create candidate variants, run those candidates inside high‑fidelity sandboxes to record runtime behavior (API calls, persistence tricks, network and file activity), convert traces into compact behavioral features, then use clustering, rules mining or machine learning to produce detectors or lightweight signatures that can be deployed to endpoints or network controls.

Useful strengths and important limits: this approach can increase early coverage for variants and reduce blind spots that signature‑only systems miss, but its value depends on mutation models and sandbox fidelity — poor models produce false positives or miss real attacker tactics. It can be compute‑intensive, raises containment and ethical considerations (never run generated malware on production networks), and attackers can adapt to evade predictable generation strategies.

Practical takeaways for payments teams and merchants: treat predictive models as one layer in a defense‑in‑depth strategy. Ask vendors for measurable evidence — independent tests, false‑positive rates on representative traffic, detection lead time, validation methodology, and data‑handling policies — and insist on safe sandboxing and rollback procedures. Combine any predictive capability with behavioral fraud scoring, rate limits, device fingerprinting, strong authentication, and a tested incident response plan. Marketing terms are useful shorthand; measurable outcomes and careful integration are what actually protect transactions.

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.

Edit Preview