13 CISOs to Watch in 2027

A CISO's public voice used to be a conference keynote once a year. In 2026 it became weekly. Security chiefs now explain incidents, argue about AI agents, recruit and push back on vendors in public, on LinkedIn, on X and on podcasts, and their peers read it the same day. The 13 people in this feature are the ones whose public work over the past year gives other security chiefs the most to learn from going into 2027.

This is not a popularity list. Follower counts were not used, and nobody was included for being well known. Each person here holds a current security seat, posted about the work in public more than once in the autumn of 2026, and put something more considered on the record during the year: a conference session, a podcast episode, an interview, a webinar, an article or a book. Each section below covers the seat, what the person talks about and how often, their 2026 work, and why their next year is worth following. Two of their own posts sit in each section, and nothing loads from LinkedIn or X until you ask for it.

How the 13 were chosen

The starting point was CISO Voices, CISO Tribune's record of what sitting security chiefs say in public. Every entry in it is checked against the person's current seat in our records before it runs, and every entry links to the original post or episode. The record began collecting posts in September 2026, so the post counts in this feature cover 1 September to 7 November 2026. By 7 November it held entries for 67 people.

Every person with entries dated 2026 was scored on three things.

First, public activity: how many posts and episodes we recorded for them, how recent they were, and how many different subjects they covered.

Second, thought leadership in 2026: sessions, panels, podcasts, interviews, webinars, articles and books listed on primary pages, such as a conference agenda, a podcast's own episode page or the person's employer. Every item named in this feature carries the passage it rests on, read again on its page on 9 November 2026.

Third, a current seat: a security leadership role confirmed in our records or on a primary page.

People were left out if their seat could not be confirmed as current, if we had fewer than two of their posts that can be shown in full, or if the subject of their posts made a fair profile difficult. Among those who qualified, the selection favoured breadth: the 13 work across AI software, banking, payments, blockchain, industrial technology, higher education, consulting and security vendors.

The counts below are the record's own and say nothing about reach: a post with three reactions and a post with three thousand count the same. We do not store the text of posts, so the descriptions of what each person posted are our own summaries of the subject.

Source: CISO Voices record, 1 September to 7 November 2026

Posts and episodes we recorded per person, 1 Sep to 7 Nov 2026

Label	entries
Sean Cassidy	10
Eric Freeman	9
Alex Stamos	8
Gerald Beuchelt	4
Leslie Nielsen	4
Michael Coates	4
Natalia Oropeza	4
Ann Barron-DiCamillo	3
Nitin Raina	3
Ty Sbano	3
Aman Sirohi	2
David Raymond	2
Mandy Andress	2

LinkedIn carried most of what we recorded. X still matters for a few of the 13, and podcasts sit in a category of their own because one episode can hold an hour of thinking.

Source: CISO Voices record, 1 September to 7 November 2026

Where the 13 posted, by channel

Label	entries
LinkedIn	45
X	12
Podcast	1

Source: CISO Voices record, 1 September to 7 November 2026 · The record began collecting posts in September 2026.

Entries we recorded for the 13, by month

Label	entries
Sep 2026	6
Oct 2026	40
1 to 7 Nov 2026	12

Ten of the 13 started their current seat in 2025 or 2026, six of them in 2026, which says something about where public voices come from: a new seat is often when a security chief starts explaining their thinking in the open.

Source: CISO Voices record, 1 September to 7 November 2026 · Not dated: our records and their employers do not give a start date.

Year each of the 13 began their current seat

Label	people
2025	4
2026	6
Not dated	3

The 2026 work behind the selection was mostly spoken rather than written. Podcasts led, followed by conference panels and talks; three of the 13 also wrote or were interviewed at length.

Source: CISO Voices record, 1 September to 7 November 2026

2026 thought leadership items on record, by type

Label	items
Podcast	10
Conference panel	4
Conference talk	2
Interview	2
Article	2
Webinar	2
Event	1
Recognition	1
Book	1

The 13, A to Z

Mandy Andress

Seat Chief Information Security Officer, Elastic. Entries we recorded, 1 Sep to 7 Nov 2026 2. 2026 talks, episodes and other items on record 3.

Mandy Andress is the Chief Information Security Officer of Elastic, the search and security analytics company. Before Elastic she led information security at MassMutual and built security programmes at TiVo, Evant and Privada. Her long career gives her a view of how the job has changed across several generations of technology, and her company's products put her close to the security operations teams that use them every day.

We recorded two of her LinkedIn posts in late October and early November 2026, both about the challenges of security leadership. Her online pace is lighter than most people here, but her 2026 speaking schedule was one of the fullest. The selection rewards that: thought leadership in this feature counts sessions and episodes as much as posts. She is a good example of a security chief whose public work happens mostly in long conversations rather than short posts, and whose views are easier to follow through those conversations than through a feed.

In March she joined InnovationAus's Commercial Disco podcast to talk about cyber risk for governments, with a focus on Australia. In September she joined the TPRM Podcast to explore how AI is changing the economics, operations and strategy of security, and in October she spoke at CyberCon 2026 in Melbourne. The thread running through these appearances is cost and scale: what security can afford to do by hand once attackers automate, and where AI earns its place on the defensive side.

In 2027, Andress brings a long operating career and an international audience to the AI security debate. Much of the public conversation among security chiefs is centred on the United States, and her repeated work with Australian government and industry audiences brings a different set of regulations and priorities into view. Elastic's own position, as a company whose platform security operations teams use for detection and investigation, also means her comments on detection, data and AI are grounded in how those teams actually work. Readers looking for an operator's view rather than a forecast will find one here. Her 2026 schedule also took her to audiences in Australia as well as the United States, so her comments often compare how different governments and industries are approaching the same AI security questions. Her career also spans the transition from on-premises security to cloud and now to AI, and her comments often place new problems in that longer history, which helps separate what is genuinely new about AI risk from what is an old problem in a new form.

Follow for: security for AI built on data discipline, international and public sector perspectives.

On the record: Speaker at CyberCon 2026 in Melbourne, InnovationAus Commercial Disco episode on cyber risk for government, TPRM Podcast episode 19 on how AI changes the economics of security.

Read more: Mandy Andress on CISO Tribune, profile on CISO Spotlights

Ann Barron-DiCamillo

Seat Executive Vice President and Global Chief Information Security Officer, U.S. Bank. Entries we recorded, 1 Sep to 7 Nov 2026 3. 2026 talks, episodes and other items on record 3.

Ann Barron-DiCamillo became Executive Vice President and Global Chief Information Security Officer at U.S. Bank in June 2026. She joined from Citi, where she led cybersecurity operations globally, and her career spans public and private sector security work. A large bank's CISO seat is among the most regulated jobs in the field, and she has used her first months in it to talk about the role itself as much as about threats.

We recorded three of her LinkedIn posts between October and early November 2026. One addresses how the role of a bank CISO is changing, another women in leadership roles in financial services, and the most recent cybersecurity in the financial sector more broadly. The pace is steady rather than heavy, and the subjects line up with the interviews she gave the same month.

Her 2026 interviews carry the substance. In October U.S. Bank published a Cybersecurity Awareness Month piece with its new CISO, built around the idea that security is the ultimate team sport and the point that fundamentals still work well against attackers who now use AI. The same week, a Help Net Security interview covered collective defense, the growing reach of the CISO role, shorter incident reporting deadlines, the balance of compliance spending against risk reduction and the sharing of threat intelligence between banks. Her argument there was that success in the role comes less from expanding authority and more from partnerships and operating models that let a whole organization manage risk together. The interview also touched on what she taught students at American University about cyber risk as a shared responsibility, a sign that she sees explaining the job as part of the job. She also appeared on a panel at an American Banker event for women in banking about how AI is changing ownership and accountability for cybersecurity and fraud.

In 2027, the questions she is raising will be tested across the banking sector. Incident reporting rules, shared intelligence and the split between security and fraud are all live issues for banks, and a global CISO at a large US bank discussing them openly gives other security chiefs a reference point. Security chiefs outside banking will also find her framing useful: the idea that the security role keeps growing and no one can own all of it, and that the job is now about operating models and partnerships, applies to most large organizations. Her next year in the seat will show how far that model can be pushed inside a heavily regulated institution.

Follow for: how a global bank CISO frames the job, collective defense and threat sharing across banks.

On the record: U.S. Bank's Cybersecurity Awareness Month piece with its new CISO, Help Net Security interview on collective cyber defense and the growing CISO role, American Banker panel on cybersecurity and fraud in the AI era.

Read more: Ann Barron-DiCamillo on CISO Tribune, profile on CISO Spotlights

Gerald Beuchelt

Seat Chief Information Security Officer, Keyfactor. Entries we recorded, 1 Sep to 7 Nov 2026 4. 2026 talks, episodes and other items on record 1.

Gerald Beuchelt became Chief Information Security Officer at Keyfactor, which sells PKI and machine identity products, on 4 November 2026. He came from Acronis, where he was CISO, and his record includes security work through Sprinklr's $4 billion IPO, the LogMeIn and GetGo merger and Salesforce's acquisition of Demandware. He has run security through transactions, which is a specialised skill: due diligence, integration and the long tail of systems that come with every deal.

We recorded four of his LinkedIn posts between the end of October and early November 2026. Three deal with core security practice: vulnerability management, patch management on Linux, and identity and access management as a way to prevent breaches. The fourth announces his move to Keyfactor. The posts are short and practical, and they tend to stay on the operational problems that consume most of a security team's week. In a year when much of the public talk among security chiefs was about AI, his focus on patching, vulnerability backlogs and identity hygiene is a reminder that a great deal of security work still starts with something old and known. That is not a contrarian position so much as an operator's one, and it is the background against which his views on AI should be read.

His main 2026 talk was an Acronis webinar in February, AI attacks LIVE, in which he and two colleagues generated AI-powered attacks in real time and walked through how to stop them. The format matters: showing attacks being built rather than describing them is one of the more effective ways to explain to boards and staff how quickly AI has lowered the cost of a convincing attack. The session had first drawn a full room at InfoSec in London before it was put online.

In 2027, the seat Beuchelt has just taken sits close to two larger technical shifts. Keyfactor sells trust infrastructure, the certificates and keys that machines and software use to prove who they are, and both AI agents and the coming move to post-quantum cryptography put that layer under pressure. A CISO who has to secure that kind of company while explaining machine identity to its customers will be close to two of the larger technical shifts of the next few years. His first year in the role should also show how an operator who has managed security through several deals sets up a programme at a company in a growth phase. Machine identity is a specialised area, so his posts from inside it are likely to be useful well beyond Keyfactor's customers.

Follow for: machine identity and certificates from the inside, practical vulnerability and patch management.

On the record: Keyfactor, Acronis webinar, AI attacks LIVE, on AI-powered attacks generated in real time.

Read more: Gerald Beuchelt on CISO Tribune, profile on CISO Spotlights

Sean Cassidy

Seat Chief Information Security Officer, Plaid. Entries we recorded, 1 Sep to 7 Nov 2026 10. 2026 talks, episodes and other items on record 1.

Sean Cassidy became Chief Information Security Officer at Plaid, the financial data company, in June 2026. He came from Asana, where he was Head of Security, after heading information security at Patreon and serving as CTO at DefenseStorm. Plaid connects bank accounts to financial apps, so its security programme carries consequences well beyond its own walls.

Cassidy is the most active person in this feature. We recorded ten entries for him between the start of September and the end of October 2026: nine LinkedIn posts and one podcast. His posts cover security controls for agentic AI, the future of software vulnerabilities, cybersecurity awareness, government policy and several security leadership topics. Many of them question received wisdom, which matches the theme of his main podcast appearance this year. The volume matters less than the pattern: his posts tend to arrive in clusters, such as four in the last week of September, which makes his feed read more like a working notebook than a set of announcements.

That appearance was the October episode of the TPRM Podcast, on why conventional security practice often fails. In it he challenges assumptions that security teams have built their programmes around, across third-party risk management, phishing, AI, security automation and threat modeling, and asks what happens when security processes create more friction than protection. The episode is built around a simple test: does a control reduce real risk, or does it mainly exist because every other company has one? It is a useful listen for any security chief who suspects some of their controls exist mainly because everyone else has them.

In 2027, Cassidy's posts are a place to watch established security practice being tested against the risk it is meant to reduce. His post on security controls for agentic AI, on 1 September, was one of the first entries in our record, and his argument that established practices should be judged by the risk they actually reduce applies directly to how companies will set rules for AI agents. In a regulated, high-trust business like Plaid's, that kind of first-principles thinking is tested against real constraints: auditors, partner banks and customers who expect proof rather than promises. His background helps here. Having run security at a collaboration software company and a creator platform, and served as CTO of a security company, before Plaid, he has seen the same controls work and fail in very different businesses.

Follow for: questioning received security practice, controls for agentic AI in a regulated business.

On the record: TPRM Podcast episode 20, on why conventional security practice often fails.

Read more: Sean Cassidy on CISO Tribune, profile on CISO Spotlights

Michael Coates

Seat Chief Information Security Officer, Solana Foundation. Entries we recorded, 1 Sep to 7 Nov 2026 4. 2026 talks, episodes and other items on record 1.

Michael Coates became the Solana Foundation's first Chief Information Security Officer in July 2026. He was previously CISO at Twitter, and his career has also included the CISO seats at Mozilla and CoinList and a founding partner role at Seven Hill Ventures. Moving into the security seat of a blockchain ecosystem foundation is an unusual step, and the role covers an ecosystem as much as an organization.

We recorded four of his posts in October 2026, three on X and one on LinkedIn. The strongest theme is AI-driven vulnerability discovery: how quickly AI tools now find and exploit flaws, and what that means for teams who have to fix them. He has also written about underinvestment in the security of internet-connected devices, and about practical details of AI developer tools. The posts are technical and short, and they read like notes from someone who still uses the tools he writes about. His time as an investor may explain why the economics of security, who pays to fix what, comes up in his posts.

His main 2026 thought-leadership item came shortly before his appointment. In March he appeared on the TPRM Podcast in an episode on why AI is breaking security as we know it, where he and the host discussed why traditional approaches are no longer enough. The argument connects directly to his posts: if attackers can find vulnerabilities with AI at machine speed, defensive programmes built on periodic testing and manual triage fall behind by design. He has followed the same idea through to its uncomfortable end in his posts on connected devices, where security has long been underfunded and where faster discovery of flaws mostly means a longer list of problems that no one is paid to fix. That gap between discovery and repair is likely to be one of the defining security problems of 2027, and he is already writing about it from a seat where it matters.

In 2027, Coates works where several difficult security problems meet. Blockchain ecosystems hold large sums in systems that are public by design, attract well-resourced attackers and rely on a wide community of independent developers. Add AI-assisted vulnerability discovery, which he writes about often, and his seat becomes an early test of whether security programmes can keep pace with attackers who automate. His experience building security functions at several companies also gives his comments on starting a programme a practical grounding that new security chiefs can use.

Follow for: AI-driven vulnerability discovery, building security for a public blockchain ecosystem.

On the record: Solana Compass, TPRM Podcast episode, AI is breaking security as we know it.

Read more: Michael Coates on CISO Tribune, profile on CISO Spotlights

Eric Freeman

Seat Chief Information Security Officer, WRITER. Entries we recorded, 1 Sep to 7 Nov 2026 9. 2026 talks, episodes and other items on record 2.

Eric Freeman is Chief Information Security Officer at WRITER, an AI company that has built its own large language model, Palmyra. He joined in August 2025 from Patreon, where he was Head of Security, after earlier security roles in the blockchain industry at Blockdaemon and BlockFi. Running security at a company that builds and sells its own models gives him a direct view of the problems that other CISOs are only starting to meet as they buy AI tools. A company that trains and serves its own models has to secure the training data, the model weights, the systems that serve answers to customers and the agents that act on their behalf, all at once, and usually at the speed of a growing startup.

He is one of the most active people in this feature. We recorded nine of his LinkedIn posts between early October and early November 2026, more than anyone here except Sean Cassidy. The subjects go well beyond product security. He has written about securing AI agents, about trends in vulnerability data over time, about gatekeeping in the security field, about the shared experience of people who work in security, and about the sandbox and lab providers used in security training. Several of them are about how people enter the profession and learn the work.

His 2026 thought leadership started from an unusual background. A May episode of the Cybersecurity Ecosystem Show traced his path from a restaurant kitchen to running security at an AI company, and set out the idea at the centre of his approach: that all of cybersecurity reduces to access control. The framing carries over naturally to AI agents, which are another identity that needs only the access its task requires. On 10 November he is the featured guest at an AegisAI dinner in San Francisco about building AI-native security stacks, an event built around his experience of building security at an AI company.

In 2027, two conversations that are usually separate meet in Freeman's work. One is the technical question of how to secure AI agents and the models behind them, which he works on every day. The other is how the security profession recruits, trains and treats its own people, which he writes about often. As AI changes what junior security work looks like, those two conversations will meet, and Freeman is already writing at the point where they do.

Follow for: access control as the core of AI agent security, how people enter the field.

On the record: Cybersecurity Ecosystem Show episode, from line cook to CISO, on AI and access control, AegisAI dinner in San Francisco on building AI-native security stacks.

Read more: Eric Freeman on CISO Tribune, profile on CISO Spotlights

Leslie Nielsen

Seat Chief Information Security Officer, Mimecast. Entries we recorded, 1 Sep to 7 Nov 2026 4. 2026 talks, episodes and other items on record 2.

Leslie Nielsen has been Chief Information Security Officer at Mimecast, the email and human risk security company, since September 2025. She joined from Klaviyo, where she was SVP and CISO. At Mimecast she is both the person who secures the company and a visible voice for its view that most security risk runs through human behaviour, a view she has extended in 2026 to AI agents.

We recorded four of her LinkedIn posts in late October and early November 2026, plus two podcast appearances earlier in the year. The posts cover risk-based security programme strategy, what CISOs should measure when they talk about incident disclosure rather than click rates, and employee behaviour as a factor in incidents. The common thread is measurement: what a security programme should track about how people behave, and what it should report.

Her 2026 podcast work set out that position in more depth. In January she joined a live CISO Series Podcast recording in New York on how automation can scale mistakes as fast as it scales good work. In a segment recorded around Black Hat 2026 for SC Media, she argued that treating agentic AI as a separate security category is the wrong instinct, because agents are an extension of human risk and the controls already used to manage people are the right base for managing machines. It is one of the clearer statements of a view that many security teams will need to test in 2027: that the agent problem is an identity and behaviour problem first. The practical consequence is that security teams may not need a new category of tool for every new kind of agent. They may instead need to extend the identity, access and monitoring controls they already run for employees, and to be honest about where those controls were weak to begin with.

In 2027, this argument will be tested in practice. As companies deploy agents with real access, the question of whether existing human risk controls stretch to cover them will be answered one deployment at a time. A CISO at a company built around human risk is well placed to report what holds up and what does not. Her posts on metrics are also worth following on their own: the shift from measuring clicks to measuring disclosure is one many security programmes are still making, and it changes what a board sees when it asks whether staff are a risk or a defense.

Follow for: AI agents as an extension of human risk, security metrics that measure behaviour rather than clicks.

On the record: Mimecast, CISO Series Podcast, recorded live in New York, on automation amplifying mistakes, SC Media segment recorded for Black Hat 2026: the agent is the new insider.

Read more: Leslie Nielsen on CISO Tribune, profile on CISO Spotlights

Natalia Oropeza

Seat Global Chief Cybersecurity Officer, Siemens. Entries we recorded, 1 Sep to 7 Nov 2026 4. 2026 talks, episodes and other items on record 3.

Natalia Oropeza is Global Chief Cybersecurity Officer at Siemens, where she is also Chief Diversity Officer. Siemens builds industrial automation, energy and infrastructure technology used in factories and plants around the world, so her remit reaches far beyond office IT into the operational systems that run physical processes.

We recorded four of her LinkedIn posts between mid-October and early November 2026. They cover AI and cybersecurity, zero trust, broader security themes and women in technology and belonging at work. The mix reflects her two roles: the security posts focus on industrial AI and architecture, while the others address who gets to build the technology.

Her 2026 thought leadership centred on industrial AI. At Hannover Messe in April she was part of a Siemens panel on how the company secures the industrial AI era, from factory operations to deterrence across its partner ecosystem. She followed it with writing on the Siemens blog, including a piece on that Hannover session with a Microsoft colleague about the asymmetry that defines industrial AI security, and on how short the time between disclosure and exploitation has become. She also took part in the Charter of Trust's fourth CyberTrust Talk, hosted by Siemens, on whether trust in technology should come from design or from regulation. The question that event put to its panel, whether governments can keep pace with technology and security risk without splintering into competing rulebooks, is one that every multinational security team will face in 2027. Earlier in the year, her writing also covered Siemens' industrial AI showcase at CES 2026 in Las Vegas, linking the company's product announcements to the security questions they raise.

In 2027, industrial AI is where security mistakes have physical consequences, and Oropeza's remit covers it. As AI is built into predictive maintenance, process optimization and autonomous control, the attack surface of factories and grids changes, and regulation in Europe is moving quickly to keep up. A CISO at one of the largest industrial technology companies, who also takes part in policy discussions on trust and regulation, is a direct line into how that shift is being handled. Security chiefs outside industry will still find her focus on partnership, rather than any single company's controls, relevant to their own supply chains. Her second role as Chief Diversity Officer also shapes her public voice, and her posts on belonging in technology sit alongside her security writing rather than apart from it.

Follow for: security for industrial AI and operational technology, the European policy debate on trust and regulation.

On the record: Charter of Trust 4th CyberTrust Talk, Trust by Design or Trust by Regulation, Hannover Messe 2026 panel on securing the industrial AI era, Siemens blog posts on industrial AI security, written after Hannover Fair 2026.

Read more: Natalia Oropeza on CISO Tribune, profile on CISO Spotlights

Nitin Raina

Seat Global Chief Information Security Officer, Thoughtworks. Entries we recorded, 1 Sep to 7 Nov 2026 3. 2026 talks, episodes and other items on record 1.

Nitin Raina is Global Chief Information Security Officer at Thoughtworks, the technology consultancy, and brings more than 25 years of experience in IT, security and risk. A consultancy's CISO has a particular job: protecting the firm's own systems and the client environments its people work inside, often at the same time. Thoughtworks staff build software inside client organizations, so every engagement adds identities, devices and access paths that a security team has to account for, under contracts that often set their own security terms. That makes his programme a useful reference for any company that relies on contractors and partners with access to its systems.

We recorded three of his LinkedIn posts in late October 2026. Two deal with security leadership topics, and one with a security leadership event. The posts are steady rather than frequent, and they reflect a CISO who is active in community events and peer groups more than in commentary on daily news. For a reader, that makes his feed a guide to where security chiefs are meeting and what they are discussing there, rather than a reaction to the headlines.

His 2026 recognition came from peers. He was named a finalist in the Large Corporate category of the 2026 ChicagoCISO ORBIE Awards, a regional programme that recognises security executives. Recognition of this kind is the thinnest form of thought leadership in this feature, and we note it as such: his public talks and interviews in 2026 were fewer than most people here. He earns his place on activity, a confirmed seat and the value of his particular vantage point.

That vantage point matters more in 2027. Consultancies and service providers are one of the main routes by which attackers reach their real targets, through shared credentials, client access and the tools consultants use. As AI agents start doing more of the work that consultants once did by hand, the question of how a service provider secures access across dozens of client environments becomes more pressing, not less. A global CISO at a large consultancy will meet that question early. His posts on security leadership, and his work in peer groups, are a way to see how it is being answered by someone responsible for both sides of the relationship. For security chiefs on the client side, the questions he faces are a useful checklist: which identities a partner's staff and tools use inside your systems, who approves their access, and how quickly it is removed when an engagement ends.

Follow for: security at a consultancy that works inside client environments, peer community work.

On the record: Thoughtworks, Finalist, Large Corporate, 2026 ChicagoCISO ORBIE Awards.

Read more: Nitin Raina on CISO Tribune, profile on CISO Spotlights

David Raymond

Seat Associate Vice President and Chief Information Security Officer, Virginia Tech. Entries we recorded, 1 Sep to 7 Nov 2026 2. 2026 talks, episodes and other items on record 2.

David Raymond became Associate Vice President and Chief Information Security Officer at Virginia Tech on 1 July 2026, succeeding a university security officer who had held the role since 2010. His path to the seat is unusual. He was the founding director of the Virginia Cyber Range and has served as the university's interim CIO and interim head of research computing, after teaching on the faculty at West Point. He is a retired Army officer, and the mix of teaching, operations and research shows in how he talks about security. Under his direction, the Virginia Cyber Range brought together security experts and educators from across Virginia to share teaching material and practice environments, an early example of the community-building that now runs through his work as CISO.

We recorded two of his LinkedIn posts in late October 2026. One points to a partnership with Amazon that advances AI and cybersecurity work; the other marks recognition for Virginia Tech's IT security team. Both are about the institution rather than about him, which is typical of university security chiefs, whose public voice is often tied to the work of their teams and students. The Amazon post points to the industry partnerships through which research universities increasingly support security and AI work.

His 2026 thought leadership came from the higher education security community. At the EDUCAUSE Cybersecurity and Privacy Professionals Conference in April he presented a session on how Virginia Tech reached 97% security awareness compliance without locking anyone out of their accounts. That is a practical result that most universities, and many companies, would like to copy: compliance gained through design and persuasion rather than enforcement. He is also co-author of On Cyber: Toward an Operational Art for Cyber Conflict, a book that applies the idea of operational art from military planning to cyber operations.

In 2027, his university is one of the places where security ideas get tested first and at scale. A large research university has open networks, tens of thousands of users with very different needs and limited budget, and it trains the next generation of security staff at the same time. Raymond's first full year in the seat will show what an operations-minded educator does with that mix, and whether results like the training compliance figure can be repeated in other areas. For security chiefs outside education, his work is a useful reference for running security where you cannot simply mandate behaviour.

Follow for: security awareness that works without lockouts, running security in an open university environment.

On the record: Virginia Tech News, Co-author of On Cyber: Toward an Operational Art for Cyber Conflict, EDUCAUSE Cybersecurity and Privacy Professionals Conference 2026 session on reaching 97% security awareness compliance.

Read more: David Raymond on CISO Tribune, profile on CISO Spotlights

Ty Sbano

Seat Chief Information Security Officer, Webflow. Entries we recorded, 1 Sep to 7 Nov 2026 3. 2026 talks, episodes and other items on record 2.

Ty Sbano has been Chief Information Security Officer at Webflow, the website building platform, since September 2025. He was previously CISO at Vercel, where he scaled security for a fast-growing developer platform, and he has held the CISO title three times. Both of his recent companies serve developers and designers who ship to the web, so his view of security is shaped by products that let customers publish quickly.

We recorded three of his posts in October 2026, two on LinkedIn and one on X. One announces a CISO conference session he was preparing for, one addresses a company's security commitment across its customer tiers, and one reflects on a difficult week in the security industry. The post on customer tiers touches a question many software companies avoid in public: whether the security a customer gets should depend on the plan it pays for. His posting is lighter than his speaking, which was busy in 2026.

In January he was the first guest of DataGrail's GrailCast Live, talking about what changes when AI becomes part of how teams build, ship and run things on the web. In the spring he appeared on Drata's When Trust Meets AI podcast in an episode about moving security from the department of no to the department of know. He opened that episode by asking what trust actually means, a question that sounds abstract but has concrete answers for a platform company: which security commitments it makes to which customers, how it proves them, and what happens when it falls short. The episode title sums up the approach it describes: security teams earn influence by explaining risk clearly and early, rather than by blocking work.

In 2027, the kind of company Sbano works for sits where AI meets publishing. Website and developer platforms are now where AI tools generate and publish code and content at a pace no review process was built for, and the security commitments they make to customers on different plans are increasingly part of how they compete. A CISO who has run security at two such platforms, and who talks openly about trust, privacy and the trade-offs of speed, is a good guide to how security will be sold and measured in that market. His GrailCast conversation with DataGrail's chief executive also put privacy alongside security, a pairing that matters more as AI tools process customer data on the web. Security chiefs at software companies will find his framing of trust as something a company proves, not something it claims, directly useful.

Follow for: security at web and developer platforms, trust and privacy when AI ships code and content.

On the record: PR Newswire, DataGrail's inaugural GrailCast Live on AI, privacy and the future of the web, Drata's When Trust Meets AI, spring 2026 season.

Read more: Ty Sbano on CISO Tribune, profile on CISO Spotlights

Aman Sirohi

Seat Chief Information Security Officer, Cyberhaven. Entries we recorded, 1 Sep to 7 Nov 2026 2. 2026 talks, episodes and other items on record 2.

Aman Sirohi has been Chief Information Security Officer at Cyberhaven, the data security company, since August 2025. His seat puts him in an unusual double role: he runs security for a company whose customers buy its product to protect their own data, so his programme is also a reference for theirs. In 2026 much of his public work dealt with one question that sits right on that line: how a company keeps track of where sensitive data goes once AI tools and agents start moving it.

We recorded two of his LinkedIn posts in October 2026, both about AI. One deals with how AI security and agentic AI change data security; the other with what CISOs should consider about agentic AI on Google Cloud. Two posts in five weeks is a modest pace next to the most active people in this feature, but the subjects are consistent, and they match his talks: he writes about the same problem he speaks about. That consistency is useful to a reader. Following him is less about catching a stream of reactions to the news and more about watching one argument develop over the course of a year, from conference panel to live session to post.

His 2026 speaking record is the main reason he is here. In October he joined MIT Technology Review for a live session on why data lineage, knowing where a piece of data came from and where it went next, is the key to securing AI agents. Earlier in the year, Cyberhaven's programme for RSAC 2026 included a CISO panel in which he appeared with the CISO of Headspace, the CSO and CIO of Greenlight and a partner at Costanoa Ventures to discuss what visibility means in data security. Both sessions put the same idea in front of different audiences: before a company can set policy for AI agents, it has to see what data those agents touch.

In 2027, data lineage is likely to move from a niche term to a line item in security budgets. As companies give AI agents access to documents, code and customer records, security teams will be asked to answer a simple question with evidence: where did this data go? Sirohi has spent the year arguing that the answer has to be traced, not inferred. His posts and talks are a practical way to watch that argument mature, and a reminder that the question is as much about data handling as about AI itself.

Follow for: data lineage as the basis for AI agent security, what visibility means in practice.

On the record: MIT Technology Review session on data lineage and securing AI agents, RSAC 2026 CISO panel on visibility in data security.

Read more: Aman Sirohi on CISO Tribune, profile on CISO Spotlights

Alex Stamos

Seat Chief Information Security Officer, Cognition. Entries we recorded, 1 Sep to 7 Nov 2026 8. 2026 talks, episodes and other items on record 2.

Alex Stamos, the former Facebook security chief, took the security seat at Cognition, the AI coding company, in September 2026. He arrives at a company whose product writes and runs code on behalf of engineers, which puts him at the point where AI agents and software security meet. His years running security at Facebook run through how he talks about the job.

Stamos is the most active X user in this group. We recorded eight of his posts between late September and the first days of November, all on X, more than anyone here except Eric Freeman and Sean Cassidy. They range widely. Some deal directly with the work, such as the safety and security questions raised by AI systems and a round of PHP and Node.js vulnerabilities that security teams were patching in October. Others are wider observations about technology, including the behaviour of autonomous vehicles. That mix is part of why people follow him: the security commentary arrives alongside a running view of how the technology industry behaves. His choice of channel also sets him apart. Most people in our record post mainly on LinkedIn; Stamos posts on X, where security researchers, journalists and engineers often join the same thread, so his posts tend to start conversations outside the executive circle as well as inside it.

His 2026 thought leadership was mostly spoken. In November he joined The Verge's Decoder podcast for a long conversation about the state of security in the AI era, where he was introduced as Cognition's security chief and a longtime expert in the field. In late October, a Ground Level AI conversation focused on what he called a coming dark age for cybersecurity and on the personal exposure CISOs now carry, from legal risk to insurance and the terms of their contracts. Both conversations turn on how AI is shifting the balance between attackers and defenders, and on who carries the blame when defense falls behind.

In 2027, Stamos's position matters as much as his opinions. Few security chiefs sit as close to an AI agent product as he now does, and he says in public what he thinks is going wrong. If the industry's arguments about agent permissions, generated code and the CISO's personal liability move in 2027, his posts and interviews are likely to be early places where they surface. Readers who disagree with his darker forecasts will still find the reasoning laid out in full, which is more useful than a summary of it. His comments on what CISOs should secure in their own contracts are practical career guidance for any security chief.

Follow for: AI security risk from inside an AI agent company, plain talk about the CISO's own liability.

On the record: Ground Level AI conversation on a coming security dark age and the pressure on CISOs, Guest on The Verge's Decoder podcast about security in the AI era.

Read more: Alex Stamos on CISO Tribune

What the 13 have in common

Three threads run through almost every section. The first is AI agents: whether the person sells software, runs a bank's security or teaches at a university, they spent a large part of 2026 explaining how agents change access, identity and accountability. The second is a return to fundamentals: several of the 13 argue that controls already in place for people, such as least privilege, training and clear ownership, are the right starting point for machines too. The third is openness about the job itself: what the role now covers, how it is reported to boards and how security teams get built.

Following them in 2027 is a cheap way for any security team to see these arguments develop before they reach a vendor's slide deck.