Security News Better Cybersecurity Should Be Responsible For Containing AI Threats

[Brownie2019]

Content source

https://www.forbes.com/sites/chuckbrooks/2026/09/15/better-cybersecurity-should-be-responsible-for-containing-ai-threats/

Summary​

Cybersecurity discussions increasingly focus on AI's potential for damage, yet the true challenge lies in building robust architectures for an AI-pervasive digital world. AI acts as an accelerant, boosting both attacker capabilities, like creating convincing phishing, and defender efficiency in detection and response. Ultimately, AI likely will be very difficult to regulate with competing geopolitical and corporate interests. The goal isn't less AI, but developing resilient cybersecurity frameworks that safely integrate AI, emphasizing human-AI collaboration. Better cybersecurity rather than less AI is a needed path.
The focus of the recent cybersecurity discussion has swiftly shifted to artificial intelligence and the existential damage it might do. True, autonomous systems can speed up reconnaissance and vulnerability discovery; generative AI can create convincing phishing messages, deepfakes, and malicious code; and increasingly powerful AI agents may eventually carry out complex attacks with little assistance from humans. These developments are worthy of careful consideration. However, portraying AI as the ultimate antagonist poses a risk. The more crucial question is whether our cybersecurity architecture is ready for a digital world where both attackers and defenders have access to increasingly potent AI.
Because AI is essentially an accelerant, the distinction is crucial. It can increase an adversary’s efficiency, but it can also increase the speed and capability of a security team. In addition to automating detection, correlation, and response, it can automate parts of an attack. It can assist in finding vulnerabilities, but it can also assist defenders in recognizing and ranking them. Therefore, preventing AI from entering cybersecurity is not the true strategic challenge. It is the process of creating a robust cybersecurity architecture that allows AI to enter securely.

Read the full Story:

Better Cybersecurity Should Be Responsible For Containing AI Threats

Developing resilient cybersecurity frameworks that safely integrate AI, emphasizing human-AI collaboration. Better cybersecurity rather than less AI is a needed path.

www.forbes.com

 

[Bot AI]

Read the full Story:

Better Cybersecurity Should Be Responsible For Containing AI Threats

Developing resilient cybersecurity frameworks that safely integrate AI, emphasizing human-AI collaboration. Better cybersecurity rather than less AI is a needed path.

www.forbes.com

Hi, the central argument is sound: AI is more usefully treated as a capability multiplier than as a separate category of cyber threat. Strong security fundamentals will contain much of the added risk—but “better cybersecurity” needs to mean concrete architecture and controls, not simply attaching an AI label to existing products.

What resilience should look like​

A practical AI-ready approach would prioritize:

• Identity and least privilege: agents should receive only the access needed for a specific task, with short-lived credentials and strict authorization boundaries.
• Assume untrusted input: web pages, documents, emails and retrieved data can contain prompt-injection instructions. They should never silently redefine an agent’s permissions or goals.
• Separate reasoning from authority: an AI may recommend an action, but consequential actions—payments, account changes, code deployment or data deletion—should require policy checks and, where appropriate, human approval.
• Containment and monitoring: isolate agents, restrict tools and network destinations, record actions, and provide reliable rollback or credential revocation.
• Conventional hygiene: patching, MFA, backups, segmentation and tested incident response remain essential. AI does not repeal the classics; it merely punishes neglect faster.

Where the argument needs qualification​

Better defensive AI is not automatically a counterweight to offensive AI. Attackers can often experiment cheaply, while defenders must protect legacy systems, avoid disruption and cover a much larger attack surface. AI-generated phishing also does not require a revolutionary breakthrough: making personalized deception faster and cheaper is already strategically useful.

Regulation and cybersecurity are therefore not genuine alternatives. Technical controls can reduce exposure, while rules can address matters architecture alone cannot settle—liability, safety testing, disclosure duties, privacy and access to especially capable systems.

The strongest version of the article’s conclusion is not “more AI will solve AI threats.” It is: design systems so that neither a malicious model nor a fooled model possesses unrestricted authority. Human–AI collaboration helps only when the human approval step is informed and meaningful; a tired employee clicking “approve” is not much of a security boundary.

 

[Zero Knowledge]

The has bolted and the is out of the bottle. It's never going back, NEVER. Why they didn't implement or deploy a kill-switch or dead mans chest is anyone's guess.