Rise of the machines

Share This Post

We have seen the films in which a computer comes to life and threatens humanities existence.

We have been told that whilst the fiction in the films is a long way away, there is some real-life concern regarding the development of AI.

AI is here, albeit in a basic if not adolescent form. OpenAI’s ChatGPT has really brough the AI concept into something that is a little more tangible than lots of talk about data science concepts and acronyms.

As cyber security experts, are we worried? Yes.

We know ChatGPT is an excellent composer of text that could be used to create accurate Phishing content, or a security risk if someone pastes confidential information to be rewritten.

We probably have all tried to do something a little bit naughty, asking ChatGPT to write some code or comment on political views. Fortunately, there have been several safeguards implemented to stop ChatGPT doing things that it shouldn’t.

So, we have access to an AI system, that has been learning its dataset for years, able to create programs, write text and whole multitude of things that can help us in everyday life.

Now let’s consider the opposite to a helpful, ‘good’ AI platform. An AI platform that does not have safeguards in place and has been trained in datasets that could be questionable in their accuracy, relevancy, or viewpoint.

Here we have the cyber criminal’s version of ChatGPT. The AI that helps write malicious code that evades security software, helps write Phishing text in multiple languages and in local dialect, identify vulnerabilities that can be exploited.

Here we have the worst of AI, where Governments cannot enforce safeguards, where its use is purely for illicit gain, and where the criminals can gain a significant advantage.

Is ‘bad’ AI here? The answer is yes. WormGPT and FraudGPT are 2 examples of criminally focussed AI platforms. You can be sure criminal gangs and nation-state sponsored hacking groups will be looking at how they can take advantage of AI in the future.

This is the start of how to fight adversaries that have access to advanced tools, with little or no regard to the victims of its output.

What can we do to protect ourselves? In most instances the basics of cyber security still hold true. Making sure systems are patched, 2 factor authentication, least privilege access, and so on.

What is becoming more critical is ensuring internal operating procedures to manage the infrastructure are robust and working. Time is a critical factor to ensure vulnerabilities are quickly patched, credentials are reset, or malicious emails are quarantined.

AI will enable cyber criminals and hackers to bring their wares to market quickly, more accurately and in some cases one step ahead of the defenders. As we move into this AI powered threat theatre AI defences will also improve. As always, this will be game of cat and mouse.

The irony is that if you ask ChatGPT how to set up an AI platform it will gladly help you through this process!

If you think your cyber security architecture, processes, procedures or current security tools are not up to scratch, please contact us and we can discuss further.

CALL:  01903 947780

EMAIL: hello@2tela.co.uk

2tela are human beings as opposed to AI, we had thought of running this text through ChatGPT but decided it would lose its human touch.

Request a call back from team 2tela

Fill in your details and we'll be in touch

More To Explore

Threats & Exploits

Cyber operations during conflict and war

Middle East tensions significantly increased following the UK designated terrorist group Hamas attacked Israel on 7th October. Cyber operations are increasing and is a growing threat to UK businesses.

Do You Want To Boost Your Business?

drop us a line and keep in touch

Please Login

First timer?

Learn how we help give businesses peace of mind and security

Fill in the form and we'll be in touch.