Meta Pauses AI Surveillance, Losing Access to Fable 5 Triggers Lawsuit, and Engineers Hit AI Paralysis

Marketing is hard, but I'll tell you a little secret. It doesn't have to be. Let me point something out. You're listening to a podcast right now and it's great. You love the host. You seek it out and download it. You listen to it while driving, working out, cooking, even going to the bathroom. Podcasts are a pretty close companion. And this is a podcast ad. Did I get your attention? You can reach great listeners like yourself with podcast advertising from Libsyn Ads. Choose from hundreds of top podcasts offering host endorsements or run a pre produced ad like this one across thousands of shows to reach your target audience audience in their favorite podcasts with Libsyn ads go to Libsynads.com that's L I B S Y N ads.com today. Hello everyone. Welcome to Future Ready Today, the number one podcast focused on the future of work. As you can probably tell, I am still in Las Vegas at the Wynn Hotel, heading back tomorrow. But as usual, I want to make sure we get these episodes out there. Today's Wednesday, June 24, 2026. And today there are actually three, sorry, four stories, not three that I want to get to. Story number one, meta rolling back their surveillance program on employees. Story number two, a company that just sued the US Government for losing access to Fable. Five, story three, software engineers experiencing workplace paralysis. And story number four, the hidden cost of AI rollouts is your best people so burning out your top talent. Now, before I get into the four episodes, I want to make sure that I touch on a couple things. Of course. First, if you like this kind of content, please consider rating and reviewing the show on Apple. We now have 255 reviews, almost 5 stars on Apple Podcasts. So thank you, thank you to everybody who's been doing that. It helps the show get discovered by more people just like you. If you want more content like this, check out my newsletter, futureofworknewsletter.com and lastly, if you're a chief people or chief human resource officer and you want to join a group of your peers who are shaping the future of work, check out futureofworkleaders.com our next monthly session is tomorrow. We're going to be joined by Patrick Lencioni, mega best selling famous author. I'm sure a lot of you know who he is and we're going to be talking about leading in an AI driven world. So having said that, on this Wednesday, June 24th, let's get into the first story. The first one, Meta built their big surveillance program. You might remember I talked about this a couple of weeks ago, where the idea was they were trying to train their AI systems by watching how employees worked. They tracked keystrokes, they tracked mouse clicks and movements, and. And they even took periodic screenshots of what was happening on your laptop. And the problem with this is that there were no privacy reviews that were done. Executives inside of Meta were given the opportunity to opt out of this, but regular employees were not. And so 1600 employees signed a petition demanding that this program be shut down. As you can imagine, meta told those 600 employees to take a hike, and they pressed ahead anyway. And then recently, what happened? There was a security configuration error that exposed data across 45,000 internal database tables, including full prompts and transcriptions, private conversations, performance reviews, and personal data to the entire company. So every Meta employee could potentially see every other Meta employee's private activity or that had been collected under this program. Meta classified it as a serious but not a top level incident. They paused this program, which, it's amazing that it took something like this for them to consider pausing it. And, and then they said they're investigating it and that there's no current evidence that the data was improperly accessed. Meta VP of AI Research, told staff the program would only resume when they were confident and in their data protection controls. Now, the futurist lens here. There are a few things that really bother me about this story. One is that executives have the option to opt out of this. And so if you're an executive at Meta, you can tell the company, hey, I don't want you tracking my mouse clicks. I don't want you taking screenshots of what I'm doing on my laptop. But regular employees had no say in this. And so that, I think, alone, tells you how this program was actually conceived. So somebody in a room knew this was uncomfortable enough that senior leaders would want to distance themselves from it, but they still rolled it out to everybody else anyway. And this is a big values failure. This isn't a technology failure. This is a values failure across the company. Now, if you zoom out for a second, Meta is probably not an outlier here. They just got caught in a way that made it very public. But the underlying idea behind this, which is using employee behavior as a training data set for AI systems, this is actually spreading quite widely across corporate America right now. And most employees have no idea this is happening. And most leaders have not thought through the second order consequences of what they're actually doing to their workforce here. This reminds me of Frederick Winslow Taylor. This is the concept of scientific management. It was around in the early 1900s where literally what Taylor would do is he would stand behind employees with a stopwatch and measure them to try to shave seconds off of the tasks that they were doing. They would measure every single movement that an employee made, time every task. And they would try to reduce human labor to data points that they can optimize in favor of efficiency. As you can imagine, workers in the factories hated it, and unions formed partially as a response to this. Now, the productivity gains that Taylor promised were real, but what ended up happening is that they dehumanized people, literally treating them like cogs in favor of productivity. And we are now, I think, in the AI or the digital tailor movement in that exact same situation, except the stopwatch is invisible. You know, you're not seeing the stopwatch. It's all being collected for you. The data collected is order of magnitudes more intimate, and the workers being observed are knowledge workers who signed up for a career, not a monitored experiment of being some sort of a guinea pig inside these organizations. And there's probably also a little bit of a deeper problem here that we should talk about, and that is that when employees raise these governance concerns to meta leadership, basically dismiss them. Ah, you're crazy. Stop it. We're not going to be doing that stuff. This is just for, you know, training. And we're not going to actually monitor and, you know, look at any of this stuff. Of course you are. Of course you are. This is not how you build a culture where people trust new technology. I don't trust any of these companies. Not meta, not OpenAI, not anthropic, not any of these companies. As far as I'm concerned. They're all a bunch of, I don't want to say liars, but they are. Let's go with liars. They lie. They lie nonstop to prop up their valuations. Now, I get the products that they're building genuinely do have value behind them, but just the doomerism, the non stop. It's going to take jobs. No, it's not going to take jobs. No, it's going to take jobs. No. Everything's going to be fine. We're not monitoring you, we're monitoring you. It's just a bunch of lying nonstop. The people who run PR and communications and all of these companies should be fired. They do the worst job I have ever seen from any company ever. It's, it's really atrocious. I mean, they need to hire people from the tobacco companies from back in the, you know, the 80s. And 90s, bring them in here, you know, to try to turn their story around. Now, I think the backlash year, when it fully arrives across all of corporate America, which it will, is going to be pretty significant. And what's happening at Meta is what's going to happen more broadly at other organizations if they go down this path. Story 2 and these, this kind of builds on this one a little bit. A company sued the US Government because they lost access to an AI model for two weeks. So if you remember what happened, I think it was June 8th or 9th, Anthropic released its Fable 5 model, which is the. The best model that's. That's ever been out there. It's part of their Mythos, the Mythos line. And so Fable 5 was a part of Mythos, and then a couple days later, it was gone. Basically, what ended up happening, the Trump administration's Commerce Department issued an export control directly of a directive which required Anthropic to cut off access to Fable 5 and its most powerful model, Mythos 5, for any foreign national anywhere in the world. An impossible task for Anthropic. They clearly couldn't figure that out. So what they did is they shut it off for everybody. So nobody has access to Fable 5 or to the Mythos line of products, although we're told it's going to be going to come back in a couple days. Now, I already covered the backstory here, that Amazon researchers found some workarounds that bypassed Fable 5 security protocols. There were some ties to China from a company that was an investor and a partner that I think gave $100 million to anthropic. So there was just. There was just a lot going on here. And if you want, you can listen back maybe two, three episodes where I unpack this in a lot of detail. Now, this week, a US Tech startup called Legion filed suit in federal court in Washington to challenge this directive. Now, Legion is a company that builds AI powered tools for attorneys, and its software development team includes Canadian nationals working from Canada. And so what ended up happening is when Fable 5 went dark, Legion said that it lost the core tool at the center of its product development instantaneously, with no alternative. Again, they're using the latest frontier models to build and create a product, a service to make money, and it gets shut off and they can't do anything about it. And the company described the harm as immediate, irreparable, and existential. And in court filings, it states that. I'm sorry, the court filing states that in a field defined by continued access to the most capable models. Any ground lost during this shutdown cannot be recovered. Now, the futurist lens here, the language that Legion used in their court filing, they said, quote, a field defined by continuous access to the most capable models. So essentially what they're doing, they're not arguing that losing one tool among many tools was disruptive. They're arguing that in their world of legal, where they're working with attorneys, you must always have access to the latest frontier model, the most powerful model available. Otherwise what's going to happen is you are going to fall behind in ways that you can't recover from. This is a new form of business dependency that I don't think leaders are thinking through and we don't have good language for it. And I don't think we really have good precedent for it. I mean, the closest historical analogy to this that I can even find, and you can tie it to some modern day things as well, is electricity in the 1900s, because factories that ran on electricity, they obviously had a real competitive advantage over those that didn't. But what happened is, of course electricity comes from the grid. The grid could fail. And when it did, entire production lines ended up going dark. And the solution wasn't to say, well, maybe we shouldn't use electricity. It was to build redundancy backup generators, to partner with multiple suppliers to have contingency planning. We take all of that for granted now because we've had a century to develop resilience infrastructure around a critical dependency. Now I live in California where our grid is not fantastic. I'm not going to get into California politics. We have lots of problems and issues here, but I have to have a home, a whole home generator here. I never used to have this when I lived in California. I've lived in this state pretty much my entire life through different parts of California. I have never, ever had to consider getting a, a whole home generator. But now I can't rely on the grid anymore. I can't assume that the utility companies, the energy companies are not just going to shut off power to the entire city or to my entire community to try to prevent something or high winds or this and that. So they shut everything off. And so now my family and I, we have to have these whole home backup generators as a redundancy. And we are in that same stage now of needing to build a redundancy infrastructure around AI because Legion and most companies out there, we don't have backup models. There's no contingency plan. They built their company assuming continuous access to the Latest frontier models, which, which is a rational assumption for anybody to make until the day that it's not. And the government or the vendors get some sort of a directive or they find something, or they're scared of something, but they can just shut it off. And this should make all of you out there who are deploying these AI tools across your company, building workflows, relying heavily on processes to serve customers, to make improvements, and efficiencies to analyze data, that should make all of you uncomfortable because this shutdown wasn't caused by a vendor failure, it was caused by a government directive. Now, in the future, it could be caused by a vendor failure. We don't know. And the way that a lot of companies are trying to hedge against this is they're multimodal, so they give access to Copilot, to Anthropic, to Claude. So you have multiple models that your company is using. But even in a multimodal environment like that, the context that one model has, for example, if I'm using Anthropic all the time and I have all my data and information in there, all of a sudden I lose access to the latest version of it. ChatGPT doesn't magically have access to the same context, the same information that I put in Anthropic. So yes, it's great that I have another model that I can use, but I'm not going to be able to use it for the same types of things. You know, the agents that I built in Anthropic that are now shut off don't magically transform to ChatGPT. So we don't really have a solid redundancy out there. Even the idea of being downgraded to a previous model. So what happened with Fable 5? When they shut it off, they basically said, okay, you're now getting downgraded to opus 4.8, which, I mean, okay, it's better than nothing, but it's not Fable 5. So this is a new category, I think, of geopolitical risk, of business risk. AI model access is becoming a supply chain vulnerability that didn't exist a year ago, two years ago, three years ago, and so now Legion, and I expect they're going to be the first of many companies over the coming years, are in federal court arguing for their very survival. Story number three. Business Insider published a story today. It's part of a multi week series called the Great Coding Reset that goes inside the psychological experience of software engineers living through the AI transformation in real time. And the picture is not the productivity utopia that AI vendors have been selling. So first a couple numbers, and I was actually quite shocked by these numbers. So major AI model releases climbed from 18 in 2023 to 69 major model releases in 2025. By mid-2026, another 30 have already dropped. A developer in Denmark, for example, who built a database just to track AI releases said that he was losing track of what was newest and best, and he codes with these models every day. Now. Behind these models is obviously a human experience because humans are using it. A New York software engineer profiled in this article called Danny Heyman Sorry Hamam describes his reaction to every new AI tool release, not his excitement, but as anxiety. He was quoted as saying, the first thought I get isn't, oh, this is so exciting. It's I'm behind, I have to learn this asap, so you start freaking out. A CEO of a fintech startup said the pace has made mastery feel pointless. It's almost not worth it for you to become a subject matter expert, because wait one more week and they're going to simplify it for you. A developer survey of roughly 7,000 respondents found that more than 4 in 10 say AI tools threaten their job security. So as AI agents now do the prompting themselves, developers are spending more time supervising AI outputs, what people are now calling bot sitting rather than writing code. Georgetown professor Cal Newport said waiting for models to spit out code is simply boring. An HR researcher told Business Insider the anxiety is so deep that some engineers are now seriously considering career pivots into sales or support roles. And a London business school professor summed up the core problem plainly, quote, the engineers are stuck because they are being asked to deliver innovation in business as usual mode. Now, the futurist lens here is that the software engineers are the canary in the coal mine here, and this is for the entire working population. Usually software engineers are the first to get the tools, the first to be measured on AI usage and adoption. I mean, this is where you see and hear all these conversations taking place. And they are of course the first to feel the psychological weight of working alongside these tools and these systems that get better at their core skill every few weeks or months. I think it was somebody, I believe it was from Anthropic that actually came out recently and said that these AI tools are making work lonely. So you're seeing the psychological impact already on people who use these tools nonstop. So what they're experiencing every day today is something that every knowledge worker will encounter within the next two to three years. And this is why I think this is not just A tech story. It's an everyone's story. This is a concept that you may have heard of. It's called future shock. This was created by Alvin Toffler. He coined it in 1970 to describe what happens, the orientation that happens when change arrives faster than people can absorb it. And so Alan Toffler thought it would be a general phenomenon, something society's experienced across decades. What software engineers are living through is a future shock compressed into weekly or monthly professional experience. Every model released is another version of this kind of future shock. And there's no end in sight for this. 18 major AI releases in 2023. 69 in 2025. 30 more in the first half of 2026. So if you're even going through since 2025 to today, it's over 100 model releases from the beginning of 2025 to today's date, June 24th. 100 major AI releases. How are you supposed to keep up with that? How is any company going to teach and train employees? I mean, anybody that's ever taken a learning, a training and development program, you know how long these things take to, to develop, to record, to plan, to script, to distribute, weeks or months to do a module, and now you have a hundred new AI models in a year and a half. It's not possible. It is just not possible. So I think there are two aspects of this story that are not getting enough attention. The first is what I would call the mastery trap. Humans, we get deep meaning from competence, and from mastery. We spend years trying to get good at something like chess, like what my daughter is doing, or tennis, what my daughter and son are doing, or robotics, what my son is doing. And then the skill ultimately becomes a part of who we are. And so when AI can produce functional code faster than an engineer, it doesn't just threaten the job, it threatens the identity of the software engineer. What am I doing here? And this is a much harder problem to figure out than just a skills gap. And so the resistance that some workers show isn't laziness. It's maybe, I don't know, akin to grief. In some way, they grieve the loss of a craft. And it's important to talk about this, not just kind of throw this up as a performance dashboard. Now, the second issue, and by the way, this is starting with software engineers. But, you know, you can imagine what's going to happen into any area where craft is so important. Filmmaking, music production, video editing, design, create, like all of these things make such a big right. We take pride in our Craft. It's why we like going to artisanal bakeries or coffee shops or restaurants, because you want to get food and experience things from masters of their craft. And so one of the arguments that some people are making is what happens when AI starts to remove that mastery of craft. The second aspect of this, of course, is the bot sitting problem. We automated parts of coding that were mechanical and repetitive. And what's left is reviewing AI output, which turns out to be its own kind of draining, soul sucking, monotonous type of work. Because now you're not creating anything. You're auditing something that a bot made for hours with high stakes attached. So it's high stakes, but it's boring, monotonous work. So the irony here is that you've used AI to replace some of the mundane, monotonous work that you don't like doing, only to create an entire new category of mundane, monotonous work that you now have to do. This is not what anybody signed up for. This is not why people became engineers. So nobody thought to actually take a step back and ask whether the work remaining after automation would actually be meaningful and be work that we want to do. So we optimized for output and we ignored the experience of the person sitting in front of it. I think we're going to start to see major, major backlashes against all of these AI tools and creation and a lot of deep questions around value and whether it's worth it. You're already seeing, look at the stock, the stock market, the AI companies out there are getting massacred left and right because there is a big question of value that's coming into play. Is it worth it? Are these tools really that valuable, that important, that crucial? Are they really saving us time? Are they really making us money when we account for all the other things that we have to do, including creating more soul sucking work. So I think there's going to be a huge battle going forward. Last story of the day. Building on top of that one, Fortune and HR Brew published a piece this week. This was built around a finding from WellHub's 2026 Return on well Being survey. 88% of people leaders say retaining top talent is their single biggest priority right now. If only there was a book on employee experience that helped people figure this out, which you can actually learn more about by going to8exlaws.com it's my latest book where I talk about the eight laws of employee experience and how to build a future ready organization. Again, that's 8exlaws.com so WellHub's Chief Revenue Officer said, quote, those top folks are getting asked to do more. In many cases, they're getting asked to support other people in building the skills within the team, end quote. So basically, what's happening, they're implementing the tools, they're upskilling their colleagues, they're modeling new behaviors, they're serving as internal AI champions. And because they're high performers, they say yes to everything, and they end up taking on more than what is sustainable. And so the warning that this story and the research is putting out here is that it's probably at their peril that organizations are not paying careful attention to who has the most critical skills and who is carrying the most weight inside the company. Now, the futurist lens here, every major transformation, every major change management program is always carried by a small number of unusually capable, very committed people who absorb more than their share of the uncertainty so that other people don't have to. I mean, you can think about your company right now. You can probably name a couple change champions or change evangelists who are out there not only doing their job, but also volunteering on committees to help other people, you know, buy into some of the tools and programs and changes that you're seeing. The Industrial Revolution, of course, had Foreman. The digital transformation of the 1990s had its IT pioneers who translated technology for the rest of the business. And so every big transformational and technological change has its internal change agents, change agents having a hard time speaking today. These are the people who go first. These are the people who figure things out, and then they bring everyone else along for the journey. And so what AI is doing is making this group of people smaller and smaller and smaller and loading them up with more and more weight and responsibility and pressure. And because AI is moving faster than any previous technology that we have ever seen, the learning never stops. As I've already said, the moment you figure out a tool, there's another one that's right behind it. And you have to deliver and you have to push through. You have to champion it, and you have to teach others and guide others and answer questions for others and come up with new ideas for your team and your function. It doesn't be. It doesn't end. And again, because high performers are saying yes to a lot of this stuff, the weight, the pressure just builds and builds and builds and then they crack. There's a term in mountaineering called summit fever, and this is the dangerous psychological pull towards the top. When you get towards the top of the mountain, that causes a Lot of climbers to ignore warning signs and keep pushing past the point of safe return. And a lot of organizations are running their very best and their most capable people straight into summit fever right now. And the summit is AI transformation. And the warning signs are everywhere. Burnout, exhaustion, quiet quitting, disengagement. You know, these are things that the. You end up taking this long pause as you're an employee. Take a deep breath before, you know, someone says yes to another task. You barely get a moment to relax. And the thing about high performers is they're less likely to tell you that it's too much. They push through, they find solutions, they deliver. And so when they're burning out or when they're exhausted, they don't file a complaint. They're going to go on LinkedIn and say they're open to work. They're going to take those calls with other recruiters. They're going to take calls with people who want to bring them into their company. They update their resume, and guess what? They end up leaving. And when they leave, they're going to take with them not just their skills, but the institutional knowledge of how these AI systems actually work, the relationships that ended up keeping your. Your rollout together of these AI tools, credibility, all these things are going to go out the door. So if you're an organization, you need to really understand the role that your high performers are playing and maybe try to segment and separate who are going to be the individual contributors who are doing a really good job, versus maybe the coaches, the AI coaches in your company. But it's becoming very challenging to be able to do both and to be able to do everything. So those are the top four stories of the day that I wanted to get to, and I want to leave you with a quote. This one is from Norbert Wiener or Weiner. He's a mathematician who founded Cybernetics, and he wrote back in the 1950s, quote, the world of the future will be an even more demanding struggle against the limitations of our intelligence, not a comfortable hammock in which we can lie down to be waited upon by our robot slaves. So he understood this 75 years ago that intelligent machines would demand more of us and not less. And the engineers inside organizations today are experiencing this paralysis as proof. So I hope you have a wonderful, wonderful rest of the day again. If you like this podcast, please rate and review it on Apple or on Spotify. And if you're interested in sponsoring the show or giving me feedback, my email jacobuture Organization. Com. Thank you so much for tuning in and watching and listening. I'll be back tomorrow.