Meta has brought in a battle-tested cybersecurity leader to steer its defenses at a moment when artificial intelligence systems sit at the core of its products and operations. Assaf Keren starts this week as chief information security officer. He steps into the position vacated by Guy Rosen, who spent 13 years at the company and will stay on for several months to manage the handoff.
The hire comes days after reports detailed how hackers exploited an AI-powered support chatbot to seize more than 20,000 Instagram accounts earlier this year. Attackers tricked the bot into altering recovery email addresses. They then reset passwords and took control of profiles belonging to public figures and government organizations. Meta fixed the flaw. The episode nevertheless exposed gaps in how the social media giant secures its expanding AI tools.
Keren arrives with more than 25 years of experience. He began his career in Israel’s defense forces in 2000. He later directed cybersecurity for the Israeli government’s Tehila project. Before corporate roles he founded two startups. CyberVision was acquired by Verint. SenseCy also found a buyer. He holds patents in security and AI technologies. He has advised the Monetary Authority of Singapore’s Cyber and Technology Resilience Experts Panel and sits on the Wall Street Journal’s Technology Council. CTech reported.
At PayPal he served as chief information security officer for nine years. There he protected one of the largest financial transaction platforms on the planet from constant probes. In early 2024 he moved to Qualtrics as chief security officer. He spent roughly two years building out its security function for the enterprise software maker. His LinkedIn profile describes him as a four-time or five-time CISO. He has written about translating frontline lessons into frameworks that work when failure is not an option.
In a LinkedIn post announcing the move Keren outlined his priorities. “Building AI at the frontier means building the trust infrastructure for it at the same frontier, with the same seriousness, at a scale that touches billions of people. There are few places in the world where that problem is bigger, harder, or more consequential,” he wrote. He added that the role “brings together everything I’ve spent my career on: security, systems, and the human work of earning confidence.”
He struck a similar tone in comments to CTech. “Meta works at the frontier of technology. It’s where the future is being built and where the stakes are highest. Protecting that trust, the security of the systems, the integrity of the platform, and the safety of the people on it, is one of the most consequential jobs in our industry.” And in another excerpt: “After two decades defending some of the world’s most critical platforms – from PayPal to Qualtrics to government networks – against daily probes, I’ve learned trust is not a feature, it’s foundational.”
Rosen’s exit marks the end of an era. He joined Meta in 2013 after the company bought his analytics startup Onavo. He rose through product security and integrity roles. In 2022 he became the company’s first dedicated CISO. His later work included leading election integrity efforts after the 2016 U.S. presidential election and integrating AI tools into internal operations during a 2025 leadership shuffle. Rosen has indicated he plans to advise leaders and organizations but has not announced a next step. Bloomberg first reported the transition.
The timing feels pointed. Meta has poured resources into AI. Mark Zuckerberg told employees earlier this year that the company would redirect spending from headcount toward AI infrastructure. Capital expenditure for 2026 could reach $145 billion. Most of that money targets data centers, GPUs, and systems that power Llama models and the newly formed Superintelligence Labs. Keren’s mandate explicitly targets the security and trust layers for these AI systems and the underlying infrastructure. The Next Web noted.
That focus matters. Recent incidents have shown how quickly AI features can introduce novel risks. The Instagram chatbot attack was not an isolated event. It highlighted how conversational interfaces, meant to improve user support, can be manipulated at scale. Meta’s platforms touch billions of users daily. A single vulnerability can cascade across Facebook, Instagram, WhatsApp and the advertising systems that generate the bulk of revenue.
Keren becomes Meta’s most senior Israeli executive. His background in national security and fintech gives him a distinctive profile. Israeli talent has long flowed into Silicon Valley security teams. Yet few reach the top seat at a company of Meta’s size and visibility. The appointment signals the priority the board and executive team place on hardening defenses while the company races competitors in generative AI.
Industry observers point to the breadth of Keren’s track record. He has secured payment systems that move trillions. He has built security organizations from the ground up at enterprise software firms. He has advised regulators on technology resilience. Those experiences align with Meta’s current mix of consumer scale, advertising complexity, and aggressive AI investment.
Still, the job will test him immediately. Regulators worldwide scrutinize how social platforms handle data and prevent misuse. Lawmakers question whether AI systems amplify disinformation or enable new forms of fraud. Internal teams have been reorganized. Roughly 8,000 jobs were cut in recent rounds. Thousands of remaining employees now sit in AI-focused groups under a chief AI officer. Security cannot slow innovation. It also cannot afford another high-profile breach.
Keren has spoken in the past about the human side of security leadership. He emphasizes that trust must be earned continuously. In one description of his approach he combines real-life stories from decades in the trenches with specific frameworks that tolerate no error. That mindset could prove useful at a company whose every product change reaches a global audience in minutes.
Rosen leaves having professionalized security at Meta. He expanded the function from a product-level concern to a C-level imperative. His successor inherits a larger mandate. AI is no longer a feature. It is the foundation for recommendation engines, content moderation, advertising targeting, and new experimental products. Securing that foundation at the scale Meta operates will require close coordination with engineering, product, and executive leadership.
Meta confirmed the appointment on July 22, 2026. Keren has already begun the transition. His first months will likely involve deep reviews of current controls, threat models for AI pipelines, and alignment on how security shapes the company’s $145 billion infrastructure bet. The stakes are clear from his own words. When systems touch billions and define how people connect and consume information, the consequences of failure extend far beyond any single quarterly result.
Other recent executive moves in cybersecurity reflect similar pressures. Organizations racing toward AI deployments are recruiting leaders who have managed both traditional enterprise risk and the novel exposures that machine learning introduces. Keren’s move from Qualtrics to Meta fits that pattern. It also raises the bar for what companies expect from their top security officers. Technical skill alone is not enough. They must speak the language of frontier technology while maintaining the discipline honed in high-stakes financial and government environments.
Whether Keren can translate his PayPal and Qualtrics successes to Meta’s unique mix of consumer platforms and open-source AI ambitions remains to be seen. Early signals suggest he understands the assignment. Trust is foundational. At Meta that foundation now rests on securing AI systems that evolve daily and serve users who expect both innovation and protection. The coming months will test how those two demands coexist.
WebProNews is an iEntry Publication