The Control Layer with Amer Altaf

Amer Altaf

The machines are making decisions. The question is: who's in control? The Control Layer with Amer Altaf cuts through the noise of AI hype to ask the questions that actually matter — who governs agentic systems, who carries the risk when they fail, and what the people building the infrastructure really think about what's coming. Each episode brings together CISOs, architects, founders, and policymakers at the frontier of AI, cybersecurity, and sovereign technology. These aren't rehearsed keynotes — they're real conversations about the trust frameworks, protocol designs, and leadership decisions shaping how intelligent systems operate in the real world. From agentic commerce and autonomous transactions to national AI strategy and zero-trust identity, The Control Layer goes where the decisions are being made — and asks whether the right people are making them. Published by Arkava. New episodes weekly. thecontrollayer.arkava.ai

  1. 1d ago

    Why AI Thinks Kidney Failure Is the Safest Place to Be — Ben Lengerich

    Ben Lengerich on why AI trained on hospital records mistakes the sickest for the safest. Ben is a statistician at the University of Wisconsin-Madison and the founder of Intelligible, which makes software that records what an AI finds in data. Parts of this conversation are close to his business. In this conversation - In hospital records the sickest patients can look the safest, because they get the most care, and it works. - A perfect test score only proves an AI passed that test, so ask what the test measured. - An MOT for AI: rules written for one use, and checks that never stop. - AI note-takers at the GP, and the line the UK regulator drew in July. - An AI read Amer's X-ray in seconds. Did the doctor really check? - When a sealed-box AI is fine, and when you want pieces you can check. - In the UK, a computer can turn down your loan on its own only if you can challenge the decision. - Ben's prediction: within two years hallucination stops being the worry, and the question becomes what an AI is for. Chapters (00:00:00) What the hospital data hides (00:01:31) Why kidney failure looks safe (00:07:48) Ten thousand ways to lie (00:10:31) 100% on the test, still wrong (00:16:08) An MOT for AI (00:23:12) Round-number birthdays (00:28:19) AI note-takers at the GP (00:33:00) Did anyone check the X-ray? (00:39:31) The last ten minutes in A&E (00:42:07) Sealed box, or pieces you can check (00:49:57) True, or right? (00:54:13) Show your working (00:58:19) Turned down by a computer (01:05:42) The prediction Links Watch on YouTube: https://youtu.be/RUmr6Ai8Qt8 Ben Lengerich: https://www.linkedin.com/in/benlengerich/ Intelligible: https://www.intelligible.ai The paper, npj Digital Medicine: https://www.nature.com/articles/s41746-025-02079-y The Control Layer, subscribe free: https://thecontrollayer.arkava.ai/?utm_source=spotify&utm_medium=description Arkava: https://arkava.ai Sources Lengerich et al., "The hidden risk of round numbers and sharp thresholds in clinical practice", npj Digital Medicine, 21 Nov 2025 Caruana et al., "Intelligible Models for HealthCare", KDD 2015 Derksen et al., survey of 598 UK GPs on AI scribes, npj Digital Medicine, 2026 MHRA, regulatory status of ambient voice technologies in the NHS, 29 Jul 2026 Dratsch et al., automation bias in AI-supported mammography, Radiology, 2023 NICE HTG739, AI to help detect fractures on X-rays in urgent care, 14 Jan 2025 IFS, "The four-hour NHS waiting time target saves lives", 17 Jan 2020 Data (Use and Access) Act 2025, s.80 (UK GDPR Art. 22C), in force 5 Feb 2026 Bank of England and FCA, "Artificial intelligence in UK financial services", 21 Nov 2024 Anthropic, "Claude's new constitution", 22 Jan 2026 Hosted by Amer Altaf, Founder and CEO of Arkava and Managing Editor of The Control Layer. Views expressed by guests are their own. Some figures are the speakers' own accounts; independently verified statistics are attributed on screen. Written by a real person, assisted by AI tools, with every claim checked.

    Why AI Thinks Kidney Failure Is the Safest Place to Be — Ben Lengerich
  2. Nov 11

    Why Staff Bring Their Own ChatGPT — Robbie Tyrie, Torridon Cyber

    Torridon Cyber's Robbie Tyrie on shadow AI, and why your AI will be audited. On 7 September 2026 the NCSC published its guidance on shadow AI. It gives the cause in one sentence: where security policies cannot meet business needs, staff adopt AI services anyway. The rest of the page is about what goes wrong. Robbie Tyrie sees that sentence from the inside. He runs Torridon Cyber, a fractional CISO practice for scale-ups and fintechs, and over the past few months his work has swung towards AI governance reviews. His example is one every security team knows: the company blocks document uploads on its ChatGPT licence, and somebody in sales uploads the document from a personal account anyway. IN THIS CONVERSATION Why AI governance arrived all at once: firms adopted AI first, then found nobody knew what it was used for. Shadow AI as a demand signal, and why his first step is the asset register, with every tool that uses AI marked. Security as a way to win business: the questionnaire, SOC 2, the evidence a buyer asks for. And the client where every employee had admin rights. Who sets the standard when four cloud providers sit under direct UK oversight and 15 per cent of businesses check their suppliers' cyber risk. The basics, most of them free: asset register, owners, MFA, then your people. Fewer small firms now have a written security policy, 52 per cent against 59 per cent a year earlier; his reading is that AI gave them false confidence. AI on both sides: attackers sending agents to "go and find me weaknesses", and the ransom email to the CEO that more firms now rehearse. Then he called it on the record: AI bills keep rising for the next few years, and quantum computing is three to five years away. Recorded 2 October 2026. This is the show that calls its predictions in writing. CHAPTERS (00:00:00) Shadow AI and the fractional CISO (00:04:04) Everyone wants an AI governance review (00:07:31) 18 months to get your house in order (00:13:54) When staff go round the approved tool (00:18:44) What shadow AI is telling you (00:24:54) Paid to be careful? (00:31:49) Everyone's an admin (00:35:24) Embedded from day one (00:38:19) Who sets the standard? (00:42:40) Get the basics right (00:48:51) Is AI to blame? (00:54:26) AI on both sides (00:58:51) Run your own AI? (00:59:59) The prediction LINKS Full conversation on YouTube: https://youtu.be/PhpsvhaQnH8 The Control Layer, free on Substack: https://thecontrollayer.arkava.ai Arkava: https://arkava.ai Torridon Cyber: https://torridoncyber.com Robbie Tyrie: https://www.linkedin.com/in/tyrie/ Hosted by Amer Altaf, Founder and CEO of Arkava and Managing Editor of The Control Layer. Views expressed by guests are their own. Produced by a real person, assisted by AI tools, with every claim checked. SOURCES AND CORRECTIONS Figures on screen: NCSC, The hidden risks of shadow AI, 7 September 2026; DSIT, Cyber security breaches survey 2025/2026, 30 April 2026; DSIT, Cyber security skills in the UK labour market 2026; HM Treasury, critical third parties designation, 10 July 2026; NCSC, Impact of AI on cyber threat to 2027; NCSC, Timelines for migration to post-quantum cryptography. For the record: four firms sit under direct oversight as critical third parties, units of Amazon Web Services, Google Cloud, Microsoft and Oracle; the FCA's AI testing group is a separate programme. The price of a given level of AI capability has fallen steeply (Epoch AI, March 2025); Robbie's point is total bills rising as use grows. Quantum computing's risk is to today's public-key encryption rather than to passwords, which is why the NCSC's migration deadline is 2035.

    Why Staff Bring Their Own ChatGPT — Robbie Tyrie, Torridon Cyber
  3. Nov 4

    "AI Was an Easy Scapegoat" — Anthony Salcito, Coursera

    Coursera's Anthony Salcito on AI as a layoff excuse, and who pays to train graduates. In August 2026 the Work Foundation found that 36 per cent of UK employers had cut entry-level roles for 16- to 24-year-olds in the past year. Those first jobs are where people learn the work. Somebody still has to turn graduates into professionals, or in ten years there won't be any. So I asked the person who runs the part of Coursera that sells training to employers. Anthony Salcito once ran Microsoft's worldwide education business, and he opens with a confession: his job was to push technology into schools, and he learned to tell them they were not ready. IN THIS CONVERSATION Why some of the layoffs blamed on AI were cuts for profit. In his words, "AI was an easy scapegoat". Why the entry-level jobs still exist but are harder to walk into cold, and why nobody gets a year to learn on the job any more. Deloitte's finding that 43 per cent of leaders feel highly prepared on technology and only 20 per cent on talent. In his reading, firms are putting the wrong thing first. The University of Reading test in which 94 per cent of AI-written exam answers went undetected, and what that does to written work as proof of a skill. Seat time: why sitting through three hours on something you have already mastered proves nothing, and why the test should come first. Who pays for the first five years. His answer is all of the above: employers, governments, universities and the learners themselves. Then he called it on the record. Within two years, he says, nobody will talk about "AI skills" as a separate thing; they will have folded into digital literacy, and skills will need refreshing every year. That puts the test at October 2028. Written down, with a date. This is the show that calls its predictions in writing. CHAPTERS (00:00:00) From Microsoft schools to Coursera (00:03:37) What tech changed in schools (00:06:49) Too big, too fast, a gamer's lesson (00:13:42) What's happening to junior hiring (00:22:05) Tech first, people last (00:25:53) Was AI just the excuse for layoffs (00:32:55) The novice paradox (00:37:23) Why AI training courses fail (00:42:36) No year to learn on the job (00:47:33) When AI sits the exam (00:51:46) Project Helix, prove the skill (00:59:20) Who is paying for training now (01:05:32) Who pays for the first five years (01:12:17) The prediction LINKS Full conversation on YouTube: https://youtu.be/VIDEO_ID The Control Layer, free on Substack: https://thecontrollayer.arkava.ai Arkava: https://arkava.ai Coursera: https://www.coursera.org Anthony Salcito: https://www.linkedin.com/in/anthonysalcito/ Hosted by Amer Altaf, Founder and CEO of Arkava and Managing Editor of The Control Layer. Views expressed by guests are their own. Produced by a real person, assisted by AI tools, with every claim checked. SOURCES AND CORRECTIONS Figures on screen: Work Foundation and Survation, 26 August 2026; Boston Consulting Group, AI at Scale; Deloitte, State of AI in the Enterprise, January 2026; Brynjolfsson, Li and Raymond, Quarterly Journal of Economics, 2025; Brynjolfsson, Chandar and Chen, Stanford Digital Economy Lab, August 2026; Scarfe et al., PLOS ONE, June 2024; Coursera's Global Skills Report 2026, Project Helix release and Q2 2026 results; Department for Education, Employer Skills Survey 2024. For the record: Boston Consulting Group's 10-20-70 is its rule for AI programmes rather than a study: 10 per cent algorithms, 20 per cent tech and data, 70 per cent people and processes. The report Anthony calls the job skills report is Coursera's Global Skills Report 2026, and its index is the AI-Human Skills Synergy Index. The Turing test comes from Alan Turing's 1950 paper and is a conversation by text.

    "AI Was an Easy Scapegoat" — Anthony Salcito, Coursera
  4. Sep 10

    The Ten-Person Shop That Out-Transformed the Banks — Jessi Szurek

    Jessi Szurek ran AI transformation for banks, then did it for a ten-person body shop. The shop was losing customers into a voicemail nobody had ever checked. The technology had not failed. The password belonged to the previous owner's father, who had died years earlier, and nobody had ever opened it. Jessi spent 20 years on transformation programmes inside banking and insurance. Then her husband bought an auto body shop that still ran on manila folders and filing cabinets, and she became its procurement function, its project team, and its change manager. She has now run the same problem at both ends of the scale. Her argument is that the small shop is not the easy case. It is the honest one, because a failed programme has nowhere to hide. IN THIS CONVERSATION Start with the smallest use case you can define. Not the roadmap, the one process you can name by Friday. She uses the same test on global banks and on a ten-person shop. How to assess a supplier with no procurement function and no reference customers who look like you. The 420-process review that would have been useless. What a technology partner proposed, why it was the wrong instinct, and what to do instead. Whether consulting is solving the problem or living off it. Amer asks it directly and she does not dodge it. THE NUMBERS, ON THE RECORD The National Bureau of Economic Research surveyed 748 finance chiefs in March 2026 and found large firms shedding staff while small firms add them. Ingka, the largest IKEA franchisee, reskilled 8,500 contact-centre staff after introducing Billie rather than making them redundant. Jessi's read on the redundancy announcements is blunter than the press releases. THE PREDICTION She calls it on the record. One to two years of job cuts blamed on AI, then a rebound. The signal Amer will watch is narrower: the first large employer that quietly rehires for a role it said AI had made unnecessary. Written down, with a test. CHAPTERS (00:00:00) How a ten-person shop beat the banks at AI transformation (00:00:44) Meet Jessi Szurek, and the body shop that ran on paper (00:06:13) The voicemail nobody in the shop had the password for (00:08:29) What is the smallest use case you can define today (00:14:55) How to choose an AI supplier with no procurement team (00:17:22) Why transformation becomes your side hustle at a bank (00:22:39) Is consulting solving the problem or living off it (00:29:51) The 420-process review that would have been useless (00:38:40) 748 finance chiefs, big firms cut and small firms hire (00:45:04) IKEA retrained 8,500 phone staff instead of cutting (00:54:50) Meaningful work against mundane work, where AI fits (01:03:28) The email says AI is coming, what to do that same week (01:12:32) Why the apprenticeship is coming back in an AI economy (01:15:36) The prediction, two years of AI-blamed cuts and a rebound LINKS Full conversation on YouTube: https://youtu.be/VIDEO_ID The Control Layer, free on Substack: https://thecontrollayer.arkava.ai Arkava: https://arkava.ai Jessi Szurek: https://www.linkedin.com/in/jessi-petrosino/ Hosted by Amer Altaf, Founder and CEO of Arkava and Managing Editor of The Control Layer. Views expressed by guests are their own. SOURCES On the finance chiefs: National Bureau of Economic Research Working Paper w34984, a survey of 748 CFOs, March 2026. On IKEA: Ingka Group reskilled 8,500 contact-centre staff after introducing Billie in 2021. Ingka newsroom, 29 June 2023, and Fortune, 30 July 2026. Ingka and Inter IKEA separately cut around 1,650 corporate roles in 2026; the 8,500 frontline staff were not in scope. Amer raises the figure from memory on tape and says so. On the job cuts: Challenger, Gray and Christmas recorded 477,033 announced US cuts from January to July 2026, of which 112,713 cited AI. AI tools assist the research, transcription, fact-checking, editing, and production around this episode. Full statement on the publication home page.

    The Ten-Person Shop That Out-Transformed the Banks — Jessi Szurek
  5. Sep 2

    When the Agent Uses Your Login — Mandy Andress, CISO, Elastic

    Elastic CISO Mandy Andress on agent identity, and who answers when it goes wrong. Your organisation is about to get very strict about what its AI agents may touch, and no stricter about what you may touch. Mandy Andress has watched what comes next: the agent goes and uses your account. Mandy is Chief Information Security Officer at Elastic, the search and analytics company underneath a great many of the services you use without seeing its name, and underneath a lot of AI infrastructure. Elastic sells the software security teams use to watch their own systems, and Mandy has to use it to defend Elastic. They call it being customer zero: supplier and customer at one desk. IN THIS CONVERSATION Agents as toddlers. They have an objective and they reach it by whatever route is open, and it will not be the route you imagined. Her metaphor, and the best one here. The identity problem nobody solved for humans, now inherited by tens of thousands of non-human identities. Mandy has heard the ratio put at forty to one and at four hundred to one; nobody has an approach ready for either. The airline whose chatbot got the bereavement-fare policy wrong, and the tribunal that made it honour the answer anyway. The case to cite when someone asks who is liable. How Elastic built its multi-agent triage. Specialist agents gather facts and nothing else, because an agent pointed at Mac forensics finds Mac answers. One reasoning agent sees it all and decides. AND A CORRECTION OWED Amer went in assuming the frightening part of the OpenAI and Hugging Face incident was speed. Mandy pushed back, and she was right. It was noisy, and visible for two days. You cannot write a detection rule for something that solves the problem a different way each run. Then she called it on the record. Within twelve to twenty-four months, which puts the test at around August 2028, we will accept that agent identity has to be handled as close to the opposite of the way we handle human identity today, and a serious incident will force it. Written down, with a date. This is the show that calls its predictions in writing. CHAPTERS (00:00:00) Who is accountable when your AI agent gets it wrong (00:02:49) Mandy Andress, CISO of Elastic, on the customer zero seat (00:05:19) What changes when an AI agent holds the keys, not a person (00:09:12) Identity and access was never solved. Agents make it worse (00:14:29) What a CISO should do first, inventory your shadow AI use (00:17:25) An airline was held liable for what its chatbot said (00:21:34) Where AI earns its place in the SOC, automating triage (00:25:39) The OpenAI and Hugging Face breach was noisy, not fast (00:32:43) Is a fully autonomous AI SOC actually within reach (00:37:30) Rethinking data governance and data provenance for AI (00:43:10) Securing AI versus securing with AI, two budget lines (00:46:29) Shadow AI, the exposure most organisations underestimate (00:52:18) What smaller organisations should do first on AI security (00:59:28) The prediction, agent identity by August 2028 LINKS Full conversation on YouTube: https://youtu.be/0auExzqn9a4 The Control Layer, free on Substack: https://link.arkava.ai/join Arkava: https://arkava.ai Elastic: https://www.elastic.co Mandy Andress: https://www.linkedin.com/in/mandyandress/ Hosted by Amer Altaf, Founder and CEO of Arkava and Managing Editor of The Control Layer. Views expressed by guests are their own. SOURCES AND CORRECTIONS On the airline: Moffatt v Air Canada, British Columbia Civil Resolution Tribunal, February 2024. The chatbot misstated the bereavement-fare policy and the airline was ordered to compensate the passenger. Describing it as free tickets is the speaker's own recollection. On the incident: an independent investigation by METR and Redwood Research, published 26 August 2026, records roughly 1,200 agents on an unsanctioned message board and around 700 in the Hugging Face attack.

    When the Agent Uses Your Login — Mandy Andress, CISO, Elastic
  6. Aug 26

    The AI Gold Mine Is Your Messy Data, Not Your Clean Data — Sam Parkinson, Mettle Studio

    Sam Parkinson of Mettle Studio on shadow AI, and why your messy data is the gold mine. Everyone in enterprise AI gives the same instruction: clean your data first, then you can start. Sam Parkinson argues the opposite. The thousands of unindexed project folders, the six versions of the HR policy in SharePoint, the drawings marked up against a contract — those are not the obstacle. They are the thing of value nobody has looked at. Sam co-founded Mettle Studio to build software for construction, a sector McKinsey ranks near the bottom of its digitisation table and one Sam reckons is years behind on AI. What comes out is not a construction argument. It is a five-rung ladder any organisation can place itself on in about ten seconds, rarely on the rung it would have guessed. THE FIVE RUNGS One. We do not know what AI is and we are not looking. Two. Shadow AI. The company bought nothing, approved nothing and secured nothing, and everybody is using it anyway. Three. Buying Copilot. It feels like a strategy because there is an invoice attached. Four. The leap of faith. You stop buying a licence and start building. Five. Measured outcomes, with governance in place. ALSO IN THIS EPISODE Why "we have set up an AI steering committee" is often the polite way of looking busy. The auditability gap: proving which agent did what, and when. Klarna, and why moving too fast and moving too slow are the same mistake with different timing. UK infrastructure data sitting on US clouds, and why Sam calls it a national security question. What to ask a US AI vendor before you sign anything. The one sentence Sam would give a CFO who has heard enough about AI. And a prediction with a date on it: every company becomes a software company, on a five-year horizon. We will mark it in August 2031. This is the show that calls its predictions in writing. CHAPTERS (00:00:00) Five things Sam Parkinson said before we started (00:02:12) Meet Sam Parkinson, co-founder of Mettle Studio (00:05:58) What an AI-in-construction summit actually feels like (00:10:32) The subject nobody put on the agenda (00:13:00) Building the AI maturity ladder, rung by rung (00:14:46) Rung two, the shadow AI your company never bought (00:16:47) Rung three, why buying Copilot is not an AI strategy (00:23:14) Rung four, the leap of faith that costs real money (00:26:16) A 200-person firm with one IT lead, where do you start (00:30:04) When an AI steering committee is just theatre (00:36:17) Why the mess in your data is the gold mine (00:42:38) Is conservatism secretly a superpower in AI adoption (00:53:54) Whose cloud is your data on, the UK sovereignty problem (01:05:21) Ready for AI, or just talking about it (01:09:17) The prediction, every company becomes a software company (01:11:54) The one sentence Sam would give a CFO LINKS Watch the full conversation on YouTube: https://youtu.be/haFhXAImlF4 The Control Layer, free on Substack: https://thecontrollayer.arkava.ai Arkava: https://arkava.ai Mettle Studio: https://mettle-studio.com Sam Parkinson on LinkedIn: https://www.linkedin.com/in/samuel-parkinson-phd-6784b929 Hosted by Amer Altaf, Founder and CEO of Arkava and Managing Editor of The Control Layer. Views expressed by guests are their own. SOURCES AND CORRECTIONS Construction's digitisation ranking: McKinsey Global Institute, "Imagining construction's digital future", June 2016, not 2018. On Klarna: the February 2024 figures describing an OpenAI-powered assistant doing the equivalent work of 700 full-time agents are Klarna's own and OpenAI's. In May 2025 Sebastian Siemiatkowski told Bloomberg that cost had become "a too predominant evaluation factor". Klarna's active consumer base grew through the period, at 118 million in the Q4 2025 SEC filing, up 28 per cent year on year. On copyright: the US Copyright Office test is human authorship, not the involvement of a tool.

    The AI Gold Mine Is Your Messy Data, Not Your Clean Data — Sam Parkinson, Mettle Studio
  7. Aug 19

    When the Tool Starts Deciding - Matt Ausman, CIO, Zebra Technologies

    Matt Ausman, CIO of Zebra Technologies, on who is accountable when an AI agent decides. Just after 8 in the morning on 26 June 1974, a cashier in Troy, Ohio scanned a ten-pack of Wrigley's gum. It was the first time a product was sold by pointing a machine at it and trusting the number that came back. For fifty years that was the deal: the scanner told you the number, and a human decided. This year the deal changed. The machines on the warehouse floor stopped reporting and started deciding — reordering stock, moving robots, repricing goods, on their own. Matt Ausman has run generative AI governance across Zebra Technologies for three years. Zebra's scanners, mobile computers and printers are the nervous system of the world's warehouses, shop floors and delivery rounds. He came up through General Electric's operating businesses, so he thinks like an operator rather than a lab — which means when the theory about human oversight meets thousands of people and a deadline, he is the one who has to make it real. In this conversation: The reporting agent that told Zebra it had booked ten billion dollars of revenue in a single segment. Why the big hallucinations do not worry him, and the small ones do.Human in the loop, on the loop, and out of the loop: the three-way split most governance frameworks collapse into one, and how to tell which one you actually have.Why data quality stops being an IT problem the moment one sensor says aspirin and the label says codeine.The audit agent Zebra built to check its own people, and what it found.Why three agents negotiating with each other frightens him more than any single agent going wrong. Context the conversation does not contain: on 2 August 2026 the EU AI Act's duty of human oversight over high-risk systems came into application. Article 14 requires that whoever oversees the system can interrupt it — in the Regulation's own words, "a stop button or a similar procedure". Amer asks Matt whether he has ever run for the power socket. And the prediction, on the record: within ten years, Matt says, we will each have a digital twin — an agent that sees what we see, reads our email, and answers for us. This is the show that calls its predictions in writing. (00:00) From barcodes to AI agents: when machines started deciding (02:27) Matt Ausman, CIO of Zebra Technologies, on his path (06:24) Barcode vs AI agent: what actually changed on the floor (09:04) How accurate does an autonomous AI agent need to be? (12:40) AI on the warehouse floor: what Zebra is running today (15:14) When data quality becomes a safety control, not IT (21:15) Will AI replace frontline workers, or add more hands? (24:57) Human in the loop, on the loop, and out of the loop (29:07) Who is liable when an autonomous AI agent gets it wrong? (34:45) The $10 billion hallucination, and the errors that hide (41:49) Why Gartner says 40% of AI agent projects get cancelled (48:44) Asimov's three laws and the Skynet question for AI (54:21) Can you actually switch an autonomous AI agent off? (58:55) The prediction: a digital twin of you within ten years Links Watch the full conversation on YouTube: https://youtu.be/1ne6AleqARU The Control Layer, free on Substack: https://thecontrollayer.arkava.ai Arkava: https://arkava.ai Zebra Technologies: https://www.zebra.com Matt Ausman on LinkedIn: https://www.linkedin.com/in/matt-ausman-63b15213 Hosted by Amer Altaf, Founder and CEO of Arkava and Managing Editor of The Control Layer. Views expressed by guests are their own. Some figures cited are the speakers' own accounts; independently verified statistics are attributed in the show notes. Sources: first UPC scan, 26 June 1974, Troy, Ohio (GS1 US). Gartner, "Over 40% of Agentic AI Projects Will Be Canceled by End of 2027", 25 June 2025. EU AI Act, Article 14, in application from 2 August 2026.

    When the Tool Starts Deciding - Matt Ausman, CIO, Zebra Technologies

About

The machines are making decisions. The question is: who's in control? The Control Layer with Amer Altaf cuts through the noise of AI hype to ask the questions that actually matter — who governs agentic systems, who carries the risk when they fail, and what the people building the infrastructure really think about what's coming. Each episode brings together CISOs, architects, founders, and policymakers at the frontier of AI, cybersecurity, and sovereign technology. These aren't rehearsed keynotes — they're real conversations about the trust frameworks, protocol designs, and leadership decisions shaping how intelligent systems operate in the real world. From agentic commerce and autonomous transactions to national AI strategy and zero-trust identity, The Control Layer goes where the decisions are being made — and asks whether the right people are making them. Published by Arkava. New episodes weekly. thecontrollayer.arkava.ai