Safety policies are needed for safe AI adoption, security leaders say

Representational image of data security
(Image credit: Kingston)

  • Suyrvey finds Generative AI is being used more than ever in cybersecurity
  • Security leaders overwhelmingly prefer GenAI through platforms
  • The benefits don't yet outweigh the risks for most

New research has claimed cybersecurity leaders aren’t ready to adopt AI as a point solution just yet, with 80% of respondents preferring GenAI delivered through cybersecurity platforms.

The CrowdStrike 2024 State of AI in Cybersecurity survey found there are mixed feelings on AI in regards to safety and privacy controls, especially as Generative AI is in its infancy, and comes with significant security risks in its current form.

Top concerns for security leaders include the exposure of sensitive data to underlying Large Language Models, adversarial attacks on GenAI tools, and the tendency for hallucinations in GenAI, along with concerns about the lack of guardrails in GenAI tools and insufficient public policy regulations.

For security experts, by security experts

Crucially, security leaders are taking steps to ensure that policy frameworks are being used responsibly, and 87% of respondents have implemented new security policies, or are in the process of developing policies to govern AI adoption.

Do the risks outweigh the rewards? Not really. Whilst 39% of cybersecurity experts think that the benefits are greater than the risks, 40% think they're comparable, and 21% think the dangers outweigh the rewards.

Unsurprisingly, security workers believe GenAI should be specifically built for cybersecurity, with 76% of respondents preferring purpose-built tools over generic, one size fits all solutions.

This is reflected in the top purchase factors for IT workers, with the ability to improve an organization’s response and detection of attacks the top priority, alongside operational efficiency, and mitigating the impact of the IT skills shortage.

As generative AI is evolving, so is the cyber landscape. GenAI is increasingly being used in threat detection and automated security solutions, with most organizations optimistic about AI’s future in cybersecurity.

You might also like

Ellen Jennings-Trace
Staff Writer

Ellen has been writing for almost four years, with a focus on post-COVID policy whilst studying for BA Politics and International Relations at the University of Cardiff, followed by an MA in Political Communication. Before joining TechRadar Pro as a Junior Writer, she worked for Future Publishing’s MVC content team, working with merchants and retailers to upload content.

Read more
Avast cybersecurity
Rise of AI is causing many firms to worry about their cybersecurity
Cartoon Phishing
Hackers use GenAI to attack more frequently and effectively
An AI face in profile against a digital background.
The truth about GenAI security: your business can't afford to “wait and see”
Red padlock open on electric circuits network dark red background
AI-powered cyber threats are becoming the biggest worry for businesses everywhere
Closing the cybersecurity skills gap
AI security: establishing the first and last layer of defense
Crowdstrike logo
Will Chinese cyberespionage be more aggressive in 2025? CrowdStrike thinks so
Latest in Security
A close-up of a phone screen showing the Telegram, Signal and WhatsApp apps
Agentic AI has “profound” issues with security and privacy, Signal President says
How to prevent cyberattacks
NTT admits hackers accessed details of almost 18,000 corporate customers in cyberattack
Woman shocked by online scam, holding her credit card outside
Cybercriminals used vendor backdoor to steal almost $600,000 of Taylor Swift tickets
Woman using iMessage on iPhone
UK government guidelines remove encryption advice following Apple backdoor spat
Cryptocurrencies
Ransomware’s favorite Russian crypto exchange seized by law enforcement
Wordpress brand logo on computer screen. Man typing on the keyboard.
Thousands of WordPress sites targeted with malicious plugin backdoor attacks
Latest in News
Hospital
Major Oracle outage hits US Federal health record systems
Sam Porter cradles a baby
Death Stranding 2: On the Beach trailer confirms June release date and an even more harrowing post-apocalyptic world
AOC Agon Pro AG276FK gaming monitor tilted slightly to the side, showing the Windows desktop screen
Windows 11 users get ready for more ‘recommendations’ from Microsoft – but I’m relieved to say these suggestions might actually be useful
The black Xbox Series S 1TB on a colorful desk mat alongside a matching controller.
The next Xbox could simply be a PC in a 'TV-friendly shell' per latest rumor
Google Chromecast 2
Chromecast users are getting increasingly angry about a weird 'untrusted device' bug that blocks casting – but a fix is coming
A close-up of a phone screen showing the Telegram, Signal and WhatsApp apps
Agentic AI has “profound” issues with security and privacy, Signal President says