Thread by @techreview

In this next block of sessions at #CyberSecure, we’re learning about the latest cybersecurity technologies, including how artificial intelligence is being used to unearth vulnerabilities.

There are benefits and drawbacks to using artificial intelligence for cybersecurity. There are also a whole host of other emerging technologies that can defend against disinformation and computational propaganda. Providing an overview at #CyberSecure is @LUMKatrusya.

We’ve written a bit about how AI is being used to detect hackers.

In 2018, for instance, we reported on how Darktrace’s unsupervised-learning models can spot hackers who have already broken in. #CyberSecure technologyreview.com/2018/11/16/139…

That same year, we also wrote about how AI was becoming a hot new thing for cybersecurity—and how hackers could foil these security algorithms by targeting the data they train on and the warning flags they look for. #CyberSecure technologyreview.com/2018/08/11/141…

How is AI being used in cybersecurity?

Machine learning can improve some aspects of spam detection, for instance, Sedova says. #CyberSecure

One concern with AI and cybersecurity is that AI and machine learning draws on open source data and that data can be poisoned. #CyberSecure

The big question is: whose responsibility is it to audit AI systems? The government, the private sector, some other body? #CyberSecure

Sedova says she's getting ready to publish new research. 👀 #CyberSecure

What advice do you have about navigating the AI landscape?

What do you need to protect on your network? Having priorities here can help clarify what you need to employ. #CyberSecure

In this #CyberSecure session, @Darktrace’s Marcus Fowler is providing actionable steps for detecting, investigating, and mitigating both known and unknown attacks at their earliest stages – before the damage is done.

Recent high-profile cyberattacks against the nation’s critical infrastructure have resulted in not only financial loss and stolen data, but also supply chain disruptions and threats to human safety.

How is Darktrace using AI?

There's a security practitioner approach. Learning normal for business is an important component. #CyberSecure

The AI conversation is so much richer than it was even two years ago, Fowler says. #CyberSecure

An IT/OT security portfolio needs to be collaborative, Fowler says. #CyberSecure

The Biden administration has put into place cybersecurity frameworks, guidelines and resources, which are steps in the right direction, Fowler says.

But the hard work is just beginning. #CyberSecure

1-2-3-4-5-6. Over 80% of data breaches leverage weak or stolen passwords, making them the single biggest security risk an organization faces. #CyberSecure

In this next session with Joy Chik of Microsoft Identity, she’s exploring a world without passwords that still maintain employee privacy and confidentiality. Think biometrics.

What is zero-trust architecture?

Chik says: at the core of it, you never trust. You always verify. #CyberSecure

Assume you'll get breached. It's not about whether you will or won't anymore. #CyberSecure

Microsoft has been on a password-less journey for several years, Chik says. #CyberSecure

Another pro of password-less logins is that it creates a better user experience. #CyberSecure

Challenges for implementing password-less logins:

-Culture and mindset change
-Phased rollout. Cloud applications first, legacy applications later on.
-Recovery of password-less method if needed.

#CyberSecure

That's it for this block of sessions! Thanks for following along. #CyberSecure