As artificial intelligence becomes increasingly central to both cybersecurity defense and offensive operations, a debate is emerging over how quickly the technology should advance. A recent LinkedIn post from Asaf Wiener, CEO and co-founder of Mate Security, has drawn significant engagement by challenging the idea that slowing the development of frontier AI necessarily makes the digital environment safer.
Wiener’s post responds to a letter from Anthropic CEO Dario Amodei advocating for “pacing the frontier,” arguing that efforts to coordinate and slow the development of increasingly capable AI systems could have an unintended consequence for cybersecurity: defenders may be constrained while attackers continue using whatever capabilities are already available.
“Pace the frontier, and you pace exactly one side of this fight, the defense side,” Wiener wrote. “The answer isn’t to slow defenders down. It’s getting trusted, verifiable AI into their hands faster, with the right guardrails in place.”
The post quickly attracted responses from cybersecurity and technology professionals, with comments agreeing, challenging and expanding on the central argument. It shows comments from various individuals, including Omer Karny, Inbal Argov, Sharon Rosenman, Noam Bar-Lev, Chris Hoard, Dineshwar Sahni and Oren Saban.
A Debate That Extends Beyond Mate Security
One reason the post has resonated is that it takes a policy discussion around AI development and reframes it through the operational reality of cybersecurity. Wiener argues that attackers are not subject to the same voluntary coordination mechanisms as AI companies operating under safety commitments.
That distinction prompted several commenters to echo his concern. Omer Karny wrote, “Exactly. Thinking that everyone will pace themselves because someone said it- just won’t work…” Meanwhile, Inbal Argov argued that the focus should extend beyond the frontier models themselves to “the mechanisms around the models that make them trusted and operationalized, and hunting down the frontier-powered attackers.”
Sharon Rosenman similarly commented, “Slowing down innovation is not a good idea and will practically not happen. Great insight,” while Noam Bar-Lev offered a more direct formulation: “100%. Slowing things down gives attackers the edge they were looking for.”
The range of responses illustrates why the post has generated more than a simple promotional discussion around Mate Security. The comments focus primarily on the broader question of how organizations should prepare for AI-enabled attacks, rather than on the company itself.
From Frontier Models To Frontline Defense
Wiener expands the argument in a longer post published on Mate Security’s website. He separates what he calls the “two frontiers”: the pace at which AI capabilities advance and the pace at which attackers are already adopting the technology available today.
His argument is that defensive organizations cannot assume that attackers will wait for industry-wide agreements, safety frameworks or slower development cycles. Instead, security teams need defensive systems capable of operating at the same speed as increasingly automated attacks.
That argument connects to a broader shift underway in security operations. AI is increasingly being discussed not simply as an analytical assistant for security teams, but as a technology capable of carrying out portions of investigation, detection and response. The challenge, Wiener argues, is ensuring that these systems remain trusted, auditable and constrained by organizational policies as they become more autonomous.
Why The Post Is Generating Engagement
The reaction also highlights the way cybersecurity executives are increasingly using LinkedIn to participate directly in industry debates. Rather than announcing a product update or company milestone, Wiener used the platform to challenge a prominent AI-policy argument and connect it to a specific problem facing security teams.
Other comments reinforced the conversational nature of the discussion. Chris Hoard described the post simply as “A good read. Thanks,” while Dineshwar Sahni added a pointed observation: “…and attackers won’t write 3000+ or 6000+ essays … You got it right mate!”
As of the publicly visible LinkedIn page, the post had accumulated comments and reactions from a range of professionals. Wiener himself has more than 13,000 LinkedIn followers, giving the discussion an established professional audience from which to spread.
The engagement suggests that the question Wiener raised extends beyond one company’s product strategy. As AI capabilities continue moving into both offensive and defensive cybersecurity, organizations will increasingly have to decide not only how to govern the technology, but how quickly they need to deploy it. For Wiener, the distinction is straightforward: slowing innovation may be part of the safety conversation, but defenders also need the tools and guardrails to operate at machine speed.
John Kevin Hao is a news and feature writer covering cybersecurity, technology, and business targeted for professional audiences.