Everyone is arguing about whether AI can actually do your job.

Can it really write the code? Can it really replace the analyst? Is it genuinely capable, or is it all hype and a stock-price story? Smart people spend all day debating this question, going back on forth on LinkedIn, on the comments, even on Substack.

And I want to tell you, as gently as I can, that it is the wrong question. Please get off the debate merry-go-round and just stop asking that.

Your job does not depend on whether AI can do it. It depends on whether the person who signs your paycheck believes AI can do it. Those are two completely different things! Please pay attention to what I am about to say.

The belief is the trigger, not the technology

I’ve spent years working on technology policy. Inside the U.S. government, and inside two of the biggest tech companies building AI. Let me share my observations.

A CEO gets an idea in their head. Maybe from a conference, maybe from a board member, maybe from a competitor’s press release. The idea is simple: AI can do a chunk of what my people do, for a fraction of the cost.

They don’t run a careful study first. They don’t pilot it quietly and measure the results. They act on the belief. The layoffs go out. And the questions, if they get asked at all, get asked much later, after thousands of people have already lost their jobs and their income.

The damage happens at the speed of belief, not the speed of capability. That single distinction explains most of what’s happening in the labor market right now.

Sometimes it isn’t even a layoff, at least not right away. It start with a hiring freeze; they freeze entry-level roles, fill only what’s essential, and a whole generation of young, debt-saddled people trying to start their careers (to pay back said student debt) finds the door has been shut in their faces.

Why the Ford rehiring isn’t the win it looks like

Here is where people push back on me. “But Soribel, look at Ford. They laid people off because of AI, and then they had to rehire people. Doesn’t that prove AI can’t actually do the work? Isn’t that a happy ending for workers?” Brouhahahaha. Take that CEO! Workers unite! HAHAHAHAHA.

But let’s actually look beyond the headlines, and please let’s use our brains here.

Ford rehired only about 350 of its most senior, very specialized engineers after laying off thousands of workers across functions. That is a partial, senior-level rehire, not a broad recovery for the workforce.

So before we call this a worker comeback story, ask the obvious question nobody’s asking: what happened to everyone else? The HR teams. The analysts. The project managers. The administrative staff. The thousands who were not among the most senior engineers called back for a very specific reason. They’re still gone.

And what happened to all of their knowledge and their IP? I’m sure Ford captured all of that in their AI systems before discarding all these workers.

And my strong suspicion is that those 350 engineers weren’t brought back to rebuild the old workforce. They were brought back to train the systems meant to eventually replace them, too. To transfer what’s in their heads into a model, so that next time, the company doesn’t need them either.

That is not a comeback story for workers. That is a company buying time while finishing the job.

Subscribe now

The three governance failures hiding underneath every layoff

None of this is happening by accident. It’s happening because of three AI governance failures, stacked directly on top of each other. I’ll go into detail on all three.

The first failure is a lack of transparency. When a company says a layoff was about “AI” or “restructuring,” nobody can verify it. There’s no disclosure, no audit, no documentation the public gets to see. What actually happened inside that decision is a black box, and the people running these companies like it that way.

The second failure is a lack of accountability. Nobody has to prove anything before they pull the trigger. A leader acts on a hunch, thousands of families absorb the consequences, and if the hunch turns out to be wrong, there is no penalty. No refund. No reckoning. Just the next earnings call.

The third failure is a lack of measurement. We cannot even agree on the scale of what’s happening. Some economists insist all of these layoffs are aboout post-pandemic overhiring correcting itself. Others say it is macroeconomic conditions and geopolitical uncertainty. And those are true facts, no arguing here.

But there’s a new crop of economists now pushing back and saying, no, this is actually about AI. At least, when it comes to entry-level hiring, economists have stopped hand-wringing and admitted that yes, AI has in fact broker the first rung of the career ladder and young recent grads are suffering for it. Overall though, there is no shared, honest metric tracking what’s actually happening to workers as a collective right now. There’s a memo when these cuts happen, and then silence.

None of these three failures are about the technology itself. They are about the people deploying it, and the fact that nobody is forcing them to show their work.

Subscribe now

Why today’s broken systems are the real warning about advanced AI

Here is the part I most need you to understand, and it’s the part both the AI optimists and the doomers tend to miss.

Everything I just described is happening with AI in a relatively early form. And we already cannot govern it. We can’t see inside it, we can’t hold it accountable, we can’t even measure it.

Now go ahead and take look at what the AI companies are actually building. Go look, I’ll wait. What they’re building are not simple tools. These are autonomous systems, designed to take on more and more human work. The major labs have been fairly explicit that replacing human labor, and capturing that share of the economy, the biggest share of the economy, is the goal. That’s what the investment is for. That’s the business plan, stated out loud.

So this is not necessarily a story about a far-off robot apocalypse, although that is absolutely a concern. But right now, at this moment, it is closer to the opposite.

If we cannot make a hiring algorithm explain a single rejection today, we will not be able to control far more capable, far more autonomous systems making decisions across entire industries tomorrow. The governance failure is identical. Only the scale changes.

That is why this matters now, while these systems are still developing and there is still a window for us regular schmucks to demand answers.

What you can actually do about it

I promised no empty platitudes about upskilling, reskilling and “learning AI” whatever the heck that means. Here are four concrete things you can do today.

1. Start by looking up your own elected officials. Not Washington politicians. Not the swamp creatures. But your specific representative, your senators, your state legislators. Have they said a single thing about AI and jobs? Sponsored any bills to address this? Do they have any staffers working on this? I’m sure the answer will be no. Most have done nothing, and that silence means the only people shaping the rules are corporate lobbyists. Change what’s loud in their inbox. Right now, the AI conversation is being dominated by data centers, which is understandable. But I’m willing to bet once people realized the intersection of AI and jobs is a dangerous one, they’ll shift to this instead.

2. Then watch what unions are doing. Whether or not you’re in one, union contracts are where real, legally binding AI protections are being negotiated right now. Clauses requiring companies to notify workers before deploying AI, language requiring AI to support workers rather than replace them. That is what real protection looks like.

3. Break the silence, too. Talk about this plainly, with your coworkers, your neighbors, people in your field. This job market is isolating, and it turns a systemic failure into private, personal shame. Don’t let it.

4. And finally, stop internalizing it. If you have sent out hundreds of applications and heard nothing back, it is not you. It’s not your resume, and it’s not that you didn’t try hard enough. You are one person facing an automated system operating at a scale no individual can beat alone. Say it until you believe it. IT IS NOT YOU.

This is one of the most consequential issues of our generation, and it deserves more than an argument about whether the software is good enough yet.

I break the whole thing down in my new video. Watch it here on YouTube and subscribe if this is the kind of thing you want more of.

I’m Soribel Feliz, an AI governance advisor writing about how algorithmic systems shape work, opportunity, and everyday life. This is Algorithms Are Personal. If this resonated, subscribe and share it with someone who needs to hear “it’s not you” today.

I just read Mark Zuckerberg’s vision for AI, called “The Future is for Everyone” and I need to drink at least 3 gallons of water ASAP. I’m so dizzy right now.

There’s so much sugar in this I think I might need some insulin, stat.

I haven’t properly digested this 6,000+ word manifesto, but if there was ever a time when you and I need to use our critical thinking skills, this is it.

I beseech you, if you try to read this sugar-loaded piece of writing, please bring your full brain.

Zuckerberg talks about the importance of PRIVACY, TRUST, the dangers of extreme concentration of power, the need for GOVERNANCE and OVERSIGHT.

And of course, buzzwords are strewn all around: “Diversity”, “empowerment”, “For all”.

Yes, he actually said that.

Oh, and he said that superintelligence is helping his daughter bake recipes and make videos in a few hours. Yup, I’m not lying

So many contradictions in his manifesto, it looks like Llama and Deepseek had twins.

Here’s only a few.

1/ The future will be more entrepreneurial and full of abundance, for everyone. Except that, most Americans still need a job for healthcare so while we may all be hustling entrepreneurs, lots of conditions will go undiagnosed because WE HAVE NO health insurance, unless we’re willing to pay thousands of dollars for subpar healthcare. So much for abundance.

2/ Same claims from 4 years ago, that everyone will have a PhD tutor with us at all times and that we will all benefit, and will contribute, to scientific advances”

Except that AI is making healthcare more expensive, not less.

Oh, and this one:

“Everyone will have an exceptionally capable personal agent that understands you, your goals, and everything you care about. Your agent will work 24/7 on your behalf to improve your relationships, health, career, finances, home management, hobbies, and more. It will free up time for the things you enjoy, and help you accomplish more than you could otherwise. It will have strong privacy and security options so you can trust it to handle all of your personal content knowing that no one else can access your information, similar to how encryption works on WhatsApp. You’ll be able to interact with your agent through any device, including your glasses to keep you present in the moment with the people you care about.

For example, my agent flags interesting information and helps me prototype ideas. It helps keep me healthy by monitoring my sleep and then watching as I train and giving feedback. My daughter loves to bake so my agent plans personalized recipes for us to make together each weekend, orders the ingredients, and then offers suggestions as we’re baking.”

I’m not trying to be negative or pessimistic, I just want to see a good, compelling vision for AI in our future, not candy-laden wish lists that have no grounding in reality.

CEOs need to come up with a better answer than this to bring the average American to their side. So far, they have failed miserably. THIS AIN’T IT.

I don’t know man, the future is so bright that we’ll have to wear sunglasses. Preferably, Meta Rayban ones to record other people without their knowledge and consent.

#algorithmsarepersonal #aigovernance #privacy #ai

Share

I just read Mark Zuckerberg’s vision for AI, called “The Future is for Everyone” and I need to drink at least 3 gallons of water ASAP. I’m so dizzy right now.

There’s so much sugar in this I think I might need some insulin, stat.

I haven’t properly digested this 6,000+ word manifesto, but if there was ever a time when you and I need to use our critical thinking skills, this is it.

I beseech you, if you try to read this sugar-loaded piece of writing, please bring your full brain.

Zuckerberg talks about the importance of PRIVACY, TRUST, the dangers of extreme concentration of power, the need for GOVERNANCE and OVERSIGHT.

And of course, buzzwords are strewn all around: “Diversity”, “empowerment”, “For all”.

Yes, he actually said that.

Oh, and he said that superintelligence is helping his daughter bake recipes and make videos in a few hours. Yup, I’m not lying

So many contradictions in his manifesto, it looks like Llama and Deepseek had twins.

Here’s only a few.

1/ The future will be more entrepreneurial and full of abundance, for everyone. Except that, most Americans still need a job for healthcare so while we may all be hustling entrepreneurs, lots of conditions will go undiagnosed because WE HAVE NO health insurance, unless we’re willing to pay thousands of dollars for subpar healthcare. So much for abundance.

2/ Same claims from 4 years ago, that everyone will have a PhD tutor with us at all times and that we will all benefit, and will contribute, to scientific advances”

Except that AI is making healthcare more expensive, not less.

Oh, and this one:

“Everyone will have an exceptionally capable personal agent that understands you, your goals, and everything you care about. Your agent will work 24/7 on your behalf to improve your relationships, health, career, finances, home management, hobbies, and more. It will free up time for the things you enjoy, and help you accomplish more than you could otherwise. It will have strong privacy and security options so you can trust it to handle all of your personal content knowing that no one else can access your information, similar to how encryption works on WhatsApp. You’ll be able to interact with your agent through any device, including your glasses to keep you present in the moment with the people you care about.

For example, my agent flags interesting information and helps me prototype ideas. It helps keep me healthy by monitoring my sleep and then watching as I train and giving feedback. My daughter loves to bake so my agent plans personalized recipes for us to make together each weekend, orders the ingredients, and then offers suggestions as we’re baking.”

I’m not trying to be negative or pessimistic, I just want to see a good, compelling vision for AI in our future, not candy-laden wish lists that have no grounding in reality.

CEOs need to come up with a better answer than this to bring the average American to their side. So far, they have failed miserably. THIS AIN’T IT.

I don’t know man, the future is so bright that we’ll have to wear sunglasses. Preferably, Meta Rayban ones to record other people without their knowledge and consent.

#algorithmsarepersonal #aigovernance #privacy #ai

Your girl is published!!!!! She turned a Linkedin rant into an op-ed.

Ok, storytime.

Last year, I asked a founder at a pitch competition what happens to her users’ data if they experience a pregnancy loss in a state with abortion restrictions.

“We’re HIPAA compliant.”

She turned and took the next question.

That moment turned into this paper.

My op-ed is officially published in Telehealth and Medicine Today:

“When AI Becomes Care: Governance Gaps in Women’s Health Technologies.”

This was a lot of work y’all. Here’s what kept me up writing it:

We live in an appified world.

There is an app for your sleep.

There’s an app for your mood.

There is an app for your steps.

There is an app for your meals.

There’s an app for your cycle, your fertility, your pregnancy.

Each one gamifies your most intimate data into streaks, scores, and insights. Each one feels personal. Helpful. Safe.

It is not safe.

Surveillance capitalism didn’t stop at your shopping habits. It moved into your body.

Women’s health apps collect symptom logs, cycle data, mood entries, sexual activity, and location -- then share it with third-parties, including Facebook!!!! with analytics providers, and advertisers most users have never heard of.

The app infers things you never typed. Pregnancy likelihood. Mental health signals. Risk categories.

That inference gets logged. Retained. Most likely repurposed.

Most of these apps are not covered by HIPAA. That’s not a technicality. That’s a governance gap that exposes millions of people.

And in a post-Dobbs world, the stakes got higher.

A missed period logged in October. A location ping near a clinic in November.

A search for “abortion pill” in December. Each data point means little alone.

Together, they build a prosecutable narrative. Women are uniquely vulnerable here because the data these apps collect can be subpoenaed, warrantied, or obtained through third-party intermediaries in states with abortion restrictions.

And the burden never falls on the company. It falls on the user.

On the teenage girl who just left home for college and got her first boyfriend.

On the single mom of 3 who just got into a new relationship.

AI governance is women’s health policy.

We just haven’t named it that yet. And until we do, the people bearing the consequences of weak governance will continue to be the ones who had the least say in building these systems.

Click on the link below to read it. Share it. Ask your app what it knows about you.

https://telehealthandmedicinetoday.com/index.php/journal/article/view/683/1674

And remember, sometimes a rant can turn into an op-ed 😉 🫠

Oh, and Algorithms are Personal™

Oh, and AI governance is not abstract!!!

#femtech #womenshealth #aigovernance

Back from Germany and have two great milestones to celebrate today. The second one is big. I need you to help me spread the word.

First, my LinkedIn Learning course just crossed 15,000+ learners!!!

I started this course because I was tired of watching people make the same AI mistakes at work over and over again.

15,000 people later, apparently I wasn't the only one who needed this conversation.

Second, and this one is big, that same course is now approved for CPE continuing education credit through NASBA.

That means you don't just learn the 8 most common mistakes people make at work with AI. You get verified, pre-approved credit toward maintaining your professional certifications while you do it.

For my #HR professionals, compliance folks, and anyone holding a certification that requires continuing education, this one is for you.

Self-paced. LinkedIn Learning. NASBA approved. You complete the videos, pass the final exam with a 70% or higher, and download your CPE certificate.

Share

The Things I Do for AI Governance

A short story about 24 hours of travel, 90 minutes of pure energy, and why Bremen, Germany changed how I think about this work.

I spent almost 24 hours traveling to Bremen. Door to door, through multiple time zones and layovers, nearly a full day of my life in airports, planes, and transit. By any reasonable measure, I should have arrived exhausted, disoriented, and running on fumes.

Instead, I walked into Constructor University and felt completely at home. That’s what happens when you’re doing work you actually believe in.

What I went to Germany to do

On July 1st, I spoke at the 2026 International Conference of Young Scientists, hosted by the Global Young Academy at Constructor University in Bremen.

The panel was called “AI Governance for a Shared Future: Science, Policy, and Global Equity.”

I meant every word of that title when I agreed to participate. And I meant every word of it while I was sitting on that stage.

I was joined by Fadi Daou, Executive Director of Globethics, and Emilija Stojmenova Duh, Associate Professor at the Faculty of Electrical Engineering at the University of Ljubljana, Slovenia. The panel was moderated by my girl, Dr. Luz Milbeth Cumba García, Founder of SciPolicy Global Strategies and Global Young Academy member.

The session was designed to explore responsible governance of machine intelligence in science and research, across world regions and disciplines. It examined emerging global frameworks with a focus on ethical considerations and equity, and specifically on including underrepresented voices in research.

That last part was why I got on the plane.

Forget the jet lag

I want to be honest about something. I arrived in Bremen running on almost no sleep, crossing multiple time zones, carrying all the physical weight of nearly 24 hours of travel. The moment that panel started, none of that existed anymore.

We went for 90 minutes. The audience wanted to keep going. I could have kept going for 90 more minutes because this is a topic so dear to my heart that exhaustion simply doesn’t get a vote. That’s what happens when the work is real and the conversation is real and the people in the room actually want to engage with the hard questions instead of performing agreement.

Dr. Cumba García moderated brilliantly. She brought out genuinely different perspectives from each panelist and kept the audience engaged in a way that made the session feel like a conversation rather than a presentation. She deserved every moment of the standing energy in that room.

The hot takes I shared

I’m going to do a longer thought piece on the full substance of the discussion we had with Fadi and Emilija. That’s coming. The conversation was rich enough to deserve its own dedicated analysis.

For now, here are the positions I took on that stage:

• AI governance is not AI ethics. These two things are related but they are not the same. We have spent years conflating them and it has created confusion about what governance actually requires in practice.

• We also haven’t clearly defined what AI governance is. As an industry and as a global policy community, we are still working with loose definitions. That matters. You cannot govern what you haven’t defined.

• AI governance is almost always an afterthought. Companies and institutions bring it up after regulatory scrutiny appears, after a lawsuit lands, after bad publicity forces their hand. That is backwards. Governance that only activates under threat is not governance. It is crisis management.

• AI governance cannot only be discussed in the coastal regions of the United States. This one matters to me personally and I said it without apology. The conversations happening in Washington and Silicon Valley are not representative of the global reality of AI deployment. AI belongs to all of us. The governance frameworks that shape it should reflect all of us. Not just the Silicon Valley elite or the Washington policy talking heads. And that includes me, guilty as charged. I am part of the DC policy world and I know how insular those conversations can be. Going to Bremen was a deliberate choice to step outside that insularity.

Why this panel mattered

The Global Young Academy brought together scientists, researchers, and policymakers from across the world for this conference. The people in that room were not all working from the same cultural framework, the same regulatory environment, or the same assumptions about what AI governance should look like. That friction was the point.

When Fadi spoke from the perspective of Globethics and the global ethics community, he was describing a set of concerns and priorities that don’t always get airtime in American AI governance conversations. When Emilija spoke from her research and engineering background in Slovenia, she was bringing a European regulatory context and an academic rigor that challenged some of the assumptions I carry from my work in US policy and tech.

That is exactly what AI governance needs. Not echo chambers. Not rooms full of people who already agree. Rooms where the disagreements are productive and the different perspectives produce something more honest than any one voice could produce alone.

What I’m taking back with me

I came home from Bremen with more questions than answers. That is what good intellectual engagement produces.

{‘m heading back home thinking about:

• what it actually means to build AI governance frameworks that work globally, not just in the jurisdictions where the major AI companies happen to be headquarter;

• the researchers in that room who are doing work on AI deployment in contexts that American tech companies have never seriously considered. And about how much the governance conversation loses when those voices are not in the room;

• Fadi’s framing around ethics and Emilija’s framing around equity and how those two lenses produce very different priorities when you sit down to actually design a governance program; and

• how I am now more convinced than ever that AI governance is not abstract, or compliance checkbox, or a DC policy debate. It is a global conversation about who has power over the systems making decisions about people’s lives, and who gets a say in how those systems are built, deployed, and constrained. Algorithms are personal. Everywhere. Not just in the United States.

That is the work. That is why I spent 24 hours traveling to Bremen and would do it again without hesitation.

The recording of the full panel is available for anyone who wants to watch the conversation in full. I’ll link it below. And I’ll be back with the longer thought piece on the substance of what we discussed. There is a lot more to say.

For now, thank you to the Global Young Academy, to Constructor University, to Dr. Cumba García for the invitation and the moderation, to Fadi and Emilija for the conversation, and to Bremen for reminding me why this work matters beyond the borders I usually operate in.

Algorithms are personal. AI governance is not abstract. And it belongs to all of us.

Access The Recording Here

Share

If you work in AI Governance, your job responsibilities just expanded.

You need to be paying close attention to U.S. politics this year, especially primary elections leading up to the midterms in November 2026.

Do NOT come here and tell me you don’t care about politics when AI governance is the #1 issue deciding U.S. local, state and congressional elections as we speak. Not in 2028, today!!!!

“When the Democratic primary for New York’s 12th congressional district was called on Tuesday night, the result capped off one of the most expensive races of its kind in the state’s history. More than $24m poured into the Manhattan contest from tech-backed financial groups as the campaign turned into a battleground for pro- and anti-AI groups to test their influence.

Much of the spending targeted candidate Alex Bores, a member of the state assembly who sponsored an AI safety bill and subsequently became a lightning rod for the tech industry. Pro-AI political action committees (Pacs) put more than $8m into the race to oppose Bores, while industry groups supporting regulation spent more than $16m to counter the attacks.

Bores ultimately wound up in second place, losing to Michael Lasher, who had deeper ties to the Democratic party establishment. Despite the tech industry’s focus on Bores, Lasher also co-sponsored the same Raise Act AI safety bill, and similarly called for big tech to be reined in.

Exactly how the exorbitant amount of tech money shaped the race is hard to determine in a crowded primary that also included the Kennedy family scion Jack Schlossberg, who finished a distant third, and former Republican turned anti-Trump influencer George Conway. What is clearer is that NY-12 shows how the AI industry is likely to descend upon campaigns this year, as November’s midterm elections approach, and tech-backed Super Pacs – committees that can raise and spend unlimited funds in favor of or against candidates – amass hundreds of millions of dollars.

Many of the groups that spent money on Bores are well equipped to replicate their influence campaigns in other races. Leading the Future, which opposed Bores and is funded by OpenAI’s president, Greg Brockman, along with the venture capitalists Marc Andreessen and Ben Horowitz, has this year raised more than $75m. Public First Action, which is more supportive of AI regulation, meanwhile received more than $20m from Anthropic.

In NY-12, the most visible effect of this AI industry funding was a flood of often misleading attack ads that put tech, and the backlash to AI, at the center of the race.

AI Governance the #1 electoral issue that [already] decides elections. I said it in 2023 and I’ll keep saying it. Don’t call yourself an AI Governance expert if you’re not following and are on top of this.

Like my pastor says, “I know I’m right”.

Remember, Algorithms are Personal™ and AI Governance is not Abstract.

Share

Aight y’all, cat’s out the bag. I’m getting ready to travel to Germany to speak on AI Governance. My German peeps, where y’all at? Let’s meet up.

On July 1st, I’ll be speaking at the 2026 International Conference of Young Scientists, hosted by the Global Young Academy at Constructor University in Bremen.

The panel is called “AI Governance for a Shared Future: Science, Policy, and Global Equity.”

And I mean every word of that title.

We will be talking about what responsible AI governance looks like across world regions and disciplines. Not just in the coastal cities of the United States.

AI belongs to all of us, so all of us should be ‘in the room and at the table’.

Not just the Silicon Valley elite. Not just Washington policy wonks, and that includes me, guilty as charged.

Globally. With a specific focus on #ethics, equity, and making sure underrepresented voices are seated at the proverbial table where these decisions get made.

That last part matters to me personally. You know why.

I’ll be joined by Fadi Daou, Executive Director of Globethics, and Emilija Stojmenova Duh, Associate Professor at the University of Ljubljana, moderated by my girl, Luz M. Cumba García, PhD of SciPolicy Global Strategies LLC.

Access Event Here

or click below:

https://agm.globalyoungacademy.net/agm-and-conference-programme-2026/#4

I’ll be sharing some insights from the panel, but I hope you can tune in.

Remember, Algorithms are Personal™ and AI Governance is not Abstract.

Share

Another client made me cry last week. Ugly cry. I had to step away from my computer and hug my cat. But this tim

e, I decided to screenshot everything and expose him. See screenshots below.

Did you see that? How dare he make me cry like that?

Anyway, the takeaway is this: no better marketing than clients' testimonials.

Remember, Algorithms are Personal™ and AI Governance is not Abstract.

#soribelknowsherish #algorithmsarepersonal #aigovernance

I didn’t either. So I built the course I wish existed when I started.

I get the same question every week on LinkedIn: “I just became the AI person at my company. How do I actually do this job?”

Privacy analysts who got AI governance added to their plate. Cybersecurity professionals told to “figure out AI.” Compliance officers handed an AI governance portfolio with zero roadmap.

They all want to know the same thing: what do I actually do on Monday morning?

I couldn’t answer everyone individually. I post a lot on LinkedIn about AI governance, but I don’t reach everyone at the same time. And honestly, the answer is too detailed for a LinkedIn post.

So I recorded my second LinkedIn Learning course. It’s called AI Governance in Practice, and it’s exactly what the title says. The operational work. The stuff nobody teaches.

This isn’t theory. It’s lived experience.

I’ve been doing this work since before we called it AI governance.

I started at Facebook doing Trust and Safety and Algorithmic Auditing. I was on the team that built the content moderation algorithm. My job was to make sure that algorithm protected users on the platform.

This was before ChatGPT launched. We called it machine learning. Algorithmic auditing. But it was the same work. Risk assessment. Stakeholder management. Cross functional translation. Making sure automated decision systems didn’t cause harm.

Slowly it evolved into what we now call AI governance. And suddenly everyone needs it. Companies are deploying AI that makes decisions about people’s lives. Hiring. Insurance claims. Medical diagnoses. Loan approvals.

When those systems fail and companies didn’t do governance work, the consequences are real. Class action lawsuits. Reputational damage. Regulatory scrutiny. Congress knocking on your door.

That’s why this work matters. And that’s why I’m teaching it.

What the course actually covers

Crafting the table of contents was hard. I had to be ruthless about what to include. I could teach this topic forever, but LinkedIn Learning has time limits.

I forced myself to write it out in one page: if you only have one takeaway from this entire course, what are the three essential skills for doing AI governance well?

Domain expertise. Skills inventory. Stakeholder management.

Those are the foundations. Everything else builds from there.

The course covers the operational work nobody talks about at conferences. How to map stakeholders, including the ones who will quietly work around you. How to build an AI inventory when shadow AI is everywhere. How to translate between engineers and lawyers in the same meeting. How to assess risk without killing innovation. How to evaluate vendor claims when everyone says they’re compliant. How to run governance meetings that result in decisions, not more meetings.

This is not about reciting regulations. You can memorize every article of the EU AI Act and still not be able to do this job.

The common mistake I see is people thinking certifications will automatically get them a job and allow them to do the work well. That’s not the case. You need what people used to call soft skills. Cross functional communication. Collaboration. Risk management.

Those aren’t soft skills. Those are the core skills. The technical parts you can learn. The organizational politics and stakeholder dynamics take years to figure out.

Why 2026 is the year of AI governance

We’ve had three good years of having fun with LLMs and AI agents. Companies deployed first and asked questions later.

That’s changing. Fast.

2026 is the year of AI governance. Companies are realizing there’s a lot of potential for AI, but also a lot of risk. They need safeguards in place to make sure what they’re deploying doesn’t get them in trouble. They need their brand intact when they use AI.

The lawsuits are piling up. The regulatory scrutiny is intensifying. The public backlash is building.

AI governance is no longer optional. It’s survival.

And the people tasked with doing it need actual guidance. Not theory. Not philosophy. Not another framework to pin on the wall.

They need operational playbooks. That’s what this course is.

Who this course is for

AI governance is for everyone. For companies implementing AI who need to protect themselves. For employees who need to understand the systems making decisions about their work. For the public who needs to demand accountability.

An informed public that knows what AI governance is becomes an empowered public. A company that cares about AI governance is a safer company that can innovate faster because they feel protected.

This course is specifically for the people who got thrown into AI governance roles whether they wanted them or not. The ones who feel lost. The ones experiencing imposter syndrome. The ones who don’t know what to do on Monday morning.

If that’s you, this course is your playbook. Your manual. Your north star for the first 90 days and beyond.

What I hope happens

The best case scenario is that someone takes this course and comes back to tell me: “Your course helped me get through the first 90 days of this job where I felt lost and clueless. You gave me a good playbook that I used to do my job. This was very useful for me.”

If one person is helped by this course, my job here is done and this was all worth it.

I hope this course is one of the pioneer courses for helping people in this function. A foundational resource they can return to when they’re stuck. Something that helps them do their job well, be successful in AI governance, and help their companies be responsible stewards of people’s data and people’s lives.

Why I’m doing this

I wrote about some of this in a book. But people don’t read anymore.

I get individual questions about AI governance constantly. I can’t answer everyone. I can’t jump on a call with everyone who needs help.

This is how I scale myself. This is how I make a mark in a field I love. It’s work I’m already doing. Why not do it at scale in a way that reaches more people?

Teaching this course was a joy because I know it will help that one person who’s struggling right now. The person who just got handed an AI governance role and has no idea where to start.

That person exists. They’re reading this right now. And they need operational guidance, not abstract principles.

That’s what I built. The course I wish existed when I started.

The course launches soon on LinkedIn Learning

AI Governance in Practice. Based on lived experience. Built for practitioners. No theory. No philosophy. Just the operational work that actually matters.

If you’ve been thrown into AI governance and you’re wondering what to do on Monday morning, this is for you.

Algorithms are personal. AI governance is not abstract. And you’re not starting from zero.

You already have the skills. This course shows you how to connect them.

More details coming soon.

#AIgovernance #LinkedInLearning #careerdevelopment #AIcompliance #operationalexcellence #riskmanagement #stakeholdermanagement #AIstrategy #professionalskills
#algorithmsarepersonal #aigovernance

79 percent of voters are worried the government has no plan to protect workers from AI. 72 percent are concerned AI will drive down wages for people like them.

These aren’t progressive activists or tech policy wonks. These are regular voters across the political spectrum who understand what’s coming.

Here’s what makes this moment different: the American public ordinarily rejects left-wing policies like jobs guarantees or single-payer healthcare. But AI is expanding the political Overton window in real time.

The argument has shifted. It’s no longer “should the government provide a safety net?” It’s “you’re all about to lose your jobs, and the choice is either you get nothing and starve, or we do something fair.”

People don’t want to be members of the permanent underclass. And they’re starting to vote accordingly.

I’ve been saying this since 2023

When I say AI is going to be the number one electoral issue for the 2028 elections, I mean it. I planted this flag in 2023 and I will not budge.

AI may influence the 2026 midterms. But voters right now are more preoccupied with cost-of-living and affordability concerns. They’re worried about groceries and rent and gas prices.

After the midterms, after more layoffs, fewer jobs for all of us, more data centers draining local resources, more economic hardship, and God forbid more instances of AI psychosis, some fatal, voters will focus on AI governance and AI policy.

People of different demographic backgrounds, different upbringings, different sociocultural and economic beliefs will become strange bedfellows, united by this issue.

Lots of single-issue voters who will decide who to cast their ballots for based strictly on candidates’ approach to regulating AI.

The one good thing that might come from this

Americans might finally get behind the idea of a stronger social safety net for their fellow citizens. Maybe, one day, our healthcare system and insurance scheme will marginally improve.

And maybe, but I honestly have a hard time even typing this, Americans will be okay with corporations paying their fair share of taxes.

You can call me a broke futurist, a crazy AI governance cat lady, an armchair visionary. I don’t care. I said what I said.

The career ladder is breaking

Even in the most limited case, AI will break the career ladder for millions of current and future workers. The industry calls this “transitional friction.”

The Oxford economist Carl Benedikt Frey puts it plainly: “Most economists will acknowledge that technological progress can cause some adjustment problems in the short run. What is rarely noted is that the short run can be a lifetime.”

A lifetime. Not a quarter. Not a fiscal year. A human lifetime of economic displacement.

That’s not friction. That’s a fracture.

What the tech CEOs are saying in private

Alex Karp from Palantir said the quiet part out loud: “The biggest challenge to AI in this country is political unrest. If I were sitting here in private with my peers, I’d be telling them the country could blow up politically and none of us are going to make any money when the country blows up.”

The tech industry knows this is coming. They’re just hoping they can extract enough value before the backlash hits.

They won’t be able to.

The workers who are losing their jobs are organizing. The communities bearing the infrastructure costs are pushing back. The parents watching their kids suffer are demanding accountability.

And all of them vote.

Algorithms are personal. AI governance is not abstract.

Remember that. When the political realignment comes, it won’t be because people suddenly understood transformer architecture or diffusion models.

It will be because they lost their job to an algorithm. Because their energy bill doubled when the data center moved in. Because their kid developed an eating disorder from recommendation algorithms. Because they couldn’t get hired because an AI screening tool rejected them for reasons they’ll never know.

That’s when AI becomes the ballot box issue I predicted in 2023.

The receipts are coming. The backlash is building. The Overton window is shifting.

Politicians who understand this early will have an advantage. The ones who don’t will learn the hard way.

Share

Recommended readings:

• Silicon Valley Is Bracing for a Permanent Underclass. Opinion piece by Jasmine Sun: https://www.nytimes.com/2026/04/30/opinion/ai-labor-work-force-silicon-valley.html?unlocked_article_code=1.e1A.zFGe.sWGP3oHShI4x&smid=nytcore-ios-share#commentsContainer

• BEYOND CAPTURE, AI Policy Through the Veil and the Particular By and :

  Pretty Sure No One Asked

  BEYOND CAPTURE

  Foreword: This is a non‑partisan, informational analysis intended to examine facts with rigor. Any perceived opinions are the authors’ alone; not those of any past, current, or future employers. Review the references, think critically, and draw your own conclusions. Our aim is reflection, not prescription; even when the truth is uncomfortable…

  Read more

  4 months ago · 8 likes · PAULINA PÉREZ and Edwin Canizalez

Share

Read more

Mercor, a $10 billion AI startup that works with companies including OpenAI and Anthropic, confirms major data breach.

This news was reported by Abderahmane Boucetta (sharing the analysis, with permission, at the end of this article).

"Mercor just handed hackers the ultimate deepfake training kit by leaking 4 terabytes of intimate user data through basic…

Read more

I have been saying this for the last three years: AI and AI governance was always going to become a hot-button electoral issue, up there with immigration, abortion and guns.

Not eventually. Now. Today.

The Ro Khanna vs. Ethan Agarwal primary race in California is the clearest proof yet.

Silicon Valley investors are funding a primary challenger against a …

Read more

That the cohort did not launch was entirely my fault. Even though I did all the work, I didn't fully believe in it or in myself. I didn't go all in.

For example, I did not send follow-up emails justifying why people should spend their hard-earned money on my course. I didn't market it well. I didn't position myself well.

You live and learn.

I am re-launching my Breaking into AI Governance course today on Maven.
The course is open for enrollment. I've already shared it with the waitlist I've been building since last year.

I'm capping it at 12, but launching with as few as 2 people. I want it to be a small group of serious people who want accelerated entry into one of the hottest emerging careers. There is an application to be completed before enrollment.
I will not take your money if this course is not a good fit for you.

This course is not a study group for certifications. This course teaches you how to leverage your existing skills to transition into an AI governance career and all its variations.

This course is not for people who want to twist their work experience into a pretzel to say they are AI governance professionals.

This is for #Privacy, #GRC and people with deep domain expertise in a few certain fields who can credibly translate their experience and position themselves as the obvious choice for the job.

I will share my applied experience as an AI governance professional and my in-depth research of the field to help you transition into it. All my frameworks.

I fully believe in this course and myself as the instructor. See the testimonials attached (shared with permission).

The price is a steal because it's the first cohort. 6 weeks of hands-on learning with me, office hours, a cohort of like-minded peers.

I believe in this course so much I embraced Maven's refund policy.

The first free Lightning Lesson will come soon, and I'll share the one I already conducted a few weeks ago.

Because AI Governance is not abstract.
Because Algorithms are Personal.

DM me if interested. I will do 15-minute sessions with whoever wants to know if this course is worth their time and money. You can also sign up HERE.

Let's go.

#algorithmsarepersonal #aigovernance

Share

• Why People Use AI Therapy Chatbots (Even Though They’re Not Safe)

• The “Are You Dead Yet?” App and What It Says About Loneliness in 2026

  The app that made me uncomfortable.

• AI Governance in Healthcare: What Safe, Responsible Clinical AI Should Look Like Now. Why U.S. health systems, physicians, and regulators must act before AI causes preventable patient harm.

• Femtech’s Reckoning: Privacy, Power, and Protection in Health Technology with Soribel Feliz. A Term of Service Podcast Episode

There’s an app in China called “Are You Dead Yet?”

Every day, users click a button to confirm they’re still alive. If they don’t click within 24 hours, the app alerts an emergency contact to check on them.

When I first heard about it, my immediate reaction was: that’s dystopian. That’s data farming. That’s surveillance capitalism exploiting people’s loneliness.

Then I sat with it for a while.

And I realized: the app exists because people are genuinely afraid they’ll die alone and no one will notice.

Not for hours. Not for days. Maybe not for weeks.

That fear is real. And it’s growing.

The loneliness crisis nobody wants to name

We talk about mental health. We talk about depression and anxiety. We talk about social media addiction and screen time.

But we don’t talk honestly about loneliness.

Not the “I wish I had more friends” kind of loneliness.

The “I could die in my apartment and decompose for three weeks before anyone finds me” kind of loneliness.

The numbers are staggering:

In the United States, more than one-third of adults aged 45 and older feel lonely, and nearly one-fourth of adults 65 and older are considered socially isolated.

One-person households are the fastest-growing household type globally. In major cities like Tokyo, Seoul, New York, and London, 30-40% of households are single occupants.

In Japan, there’s a term for people who die alone and aren’t discovered for weeks or months: kodokushi (孤独死). It’s common enough that there are specialized companies that clean apartments after these deaths.

The U.S. Surgeon General declared loneliness a public health epidemic with mortality impacts equivalent to smoking 15 cigarettes a day.

Loneliness isn’t just an emotional state. It’s a structural crisis.

And technology isn’t causing it. Technology is responding to it.

Why “Are You Dead Yet?” exists

The app didn’t appear because someone thought “let’s exploit lonely people for profit.”

It appeared because existing social infrastructure collapsed and nothing replaced it.

What we used to have:

• Multi-generational households where family checked on elders daily

• Neighbors who noticed if you didn’t come outside

• Community spaces (churches, community centers, social clubs) where absence was noticed

• Workplaces with social bonds beyond productivity

• Dense urban neighborhoods where people knew each other

What we have now:

• People aging in place, alone

• Adult children living hundreds or thousands of miles away

• Neighborhoods where people don’t know who lives next door

• Remote work eliminating casual workplace connection

• Third places (cafes, libraries, community centers) closing or becoming transactional

• Social media giving the illusion of connection without actual support

Into this void, an app appears.

It asks: “Are you dead yet?”

And for some people, that question -however blunt- is the only daily confirmation that someone, somewhere, would notice if they were gone.

The privilege of dismissing loneliness tech

When I first shared this example on LinkedIn, some of the responses included:

“This is just a data harvesting scheme.”

“People should just call their family every day.”

“This is what happens when we let technology replace human connection.”

All technically true.

And all spoken from a place of profound privilege.

The assumption that you have family to call.

Not everyone does. Some people outlived their family. Some were estranged. Some never had close family to begin with.

The assumption that someone would notice if you stopped responding.

If you have a job, a partner, close friends, active group chats, regular social commitments—yes, someone would notice within 24-48 hours if you disappeared.

But millions of people don’t have that.

Retirees with no work structure. Divorced or widowed people whose social circles dissolved. People with disabilities who can’t leave home easily. Immigrants far from home with weak local ties.

For them, “Are You Dead Yet?” isn’t exploitative.

It’s the only thing asking.

AI companionship apps and the loneliness economy

“Are You Dead Yet?” is just one example of loneliness tech, technology designed to address social isolation.

The AI companionship industry is booming:

Replika, Character.AI, and other AI companion apps have tens of millions of users who form relationships with chatbots. Some users talk to their AI companion for hours every day.

AI boyfriends and girlfriends offer romantic connection, emotional support, and intimacy simulation for people who struggle to form human relationships.

Virtual pets and AI friends provide low-stakes social interaction for people with social anxiety or neurodivergence.

Elder care monitoring apps track activity patterns and alert family when something seems wrong—essentially automated wellness checks.

The reaction from critics is predictable:

“This is sad.”

“People should just make real friends.”

“AI companions are replacing human connection.”

Again, all true.

And again, all missing the point.

What loneliness tech reveals about what broke

AI companions didn’t replace human connection.

Human connection collapsed first, and AI companions filled the gap.

Let me be specific about what broke:

Economic precarity destroyed social infrastructure.

People move constantly for work. They can’t afford to live near family. They work multiple jobs with irregular schedules that make maintaining friendships nearly impossible.

Urban design isolated us.

Car-dependent suburbs with no walkable community spaces. Cities where you need an appointment and a credit card to exist in public. Housing costs that push people into studio apartments where they can’t host gatherings.

Digital communication replaced embodied presence.

We traded coffee dates for text threads. We traded weekly game nights for occasional likes on Instagram. We convinced ourselves that connection could happen asynchronously, at our convenience.

The pandemic accelerated everything.

Remote work normalized not seeing colleagues. Mutual aid networks dissolved once the acute crisis passed. People got used to being alone and forgot how to not be.

Aging populations with no care infrastructure.

The U.S. has no functional elder care system. Adult children are expected to be full-time caregivers while working full-time jobs. When that’s impossible, elders age in place, alone, with no daily human contact.

Into this landscape, an app appears and asks: “Are you dead yet?”

And we call it dystopian.

It is dystopian.

But the app isn’t the dystopia. The dystopia is the social infrastructure collapse that made the app necessary.

The male loneliness epidemic and algorithmic isolation

There’s a specific subset of loneliness tech I need to address: AI girlfriends and the male loneliness crisis.

Young men, particularly in their 20s and 30s, are experiencing unprecedented social isolation:

• 30% of men under 30 report having no close friends

• Male suicide rates are climbing

• Dating app algorithms optimize for engagement, not connection, leaving men feeling disposable

• Online spaces offer radicalization pipelines that promise belonging

AI girlfriends aren’t the problem. They’re a symptom.

When human connection feels impossible—when dating apps reject you, when friendships from school have dissolved, when work is remote and transactional—an AI that listens without judgment starts to feel like the only safe option.

I’m not defending this as healthy. I’m saying: we can’t dismiss it as pathetic without asking what failed these men first.

The same social infrastructure collapse that leaves elders checking “Are you dead yet?” is leaving young men forming parasocial relationships with chatbots.

Both are responses to isolation.

Both reveal that we’ve built a society where technology asking “are you okay?” feels safer than asking a human.

What we’re not building while we build loneliness tech

Here’s what makes me angry about loneliness tech:

Not that it exists.

That we’re pouring billions into technological Band-Aids instead of rebuilding the social infrastructure that prevents loneliness in the first place.

We could be building:

Community care networks: Neighborhood-based systems where people check on each other, share meals, provide mutual support.

Accessible third places: Libraries, community centers, parks, cafes designed as gathering spaces, not profit centers.

Intergenerational housing: Mixed-age communities where elders aren’t isolated and young people aren’t atomized.

Universal care infrastructure: Publicly funded elder care, childcare, disability support that doesn’t rely on family being available 24/7.

Walkable neighborhoods: Urban design that facilitates casual daily interaction instead of requiring a car to exist.

Work structures that allow for community: Reasonable hours, local jobs, schedules that permit participation in civic and social life.

Mental health support that includes loneliness: Treatment for social isolation as a public health issue, not a personal failing.

Instead, we get an app that asks if you’re dead yet.

AI governance can’t solve social infrastructure collapse

I work in AI governance. My job is to assess risk, build frameworks, ensure technology doesn’t harm people.

But AI governance has limits.

I can assess the privacy risks of “Are You Dead Yet?” I can evaluate data handling practices. I can require transparency about who has access to check-in data.

But I can’t govern away the loneliness that made the app necessary.

That requires different work:

• Policy that funds community infrastructure

• Urban planning that facilitates connection

• Economic structures that don’t require constant mobility

• Cultural shifts that value care work and community building

AI governance can make loneliness tech safer.

It can’t make it unnecessary.

The gray area of compassion

Here’s where I land on “Are You Dead Yet?” and similar loneliness tech:

It’s exploitative. Someone is profiting from isolation.

It’s dystopian. The fact that it’s needed reveals societal failure.

And it’s compassionate. For some people, it’s the only thing checking.

All three can be true.

The person using “Are You Dead Yet?” isn’t stupid or pathetic. They’re responding rationally to their circumstances.

The developer who built it isn’t necessarily evil. They might have seen a real need and built a real solution.

The investors funding loneliness tech aren’t all predatory vultures. Some genuinely believe they’re solving a problem.

And the society that made this app necessary is all of us.

We chose urban sprawl over walkable neighborhoods.

We chose efficiency over community.

We chose economic growth over social infrastructure.

We chose to let elder care become a private family problem instead of a public responsibility.

An app asking “are you dead yet?” is what happens when those choices compound.

What comes next

I don’t have solutions for the loneliness epidemic.

I have observations:

Technology will keep filling the gaps we create. If we don’t rebuild community infrastructure, AI companions will become more sophisticated, more emotionally compelling, more necessary.

Loneliness is profitable. As long as isolation grows, someone will monetize the solution. The question is whether we regulate that market or let it operate unchecked.

Dismissing loneliness tech as dystopian doesn’t help the people using it. They already know it’s not ideal. They’re using it anyway because the alternative is worse.

We can’t app our way out of social infrastructure collapse. At some point, we need to decide: do we want a society where humans care for each other, or a society where we outsource that to algorithms?

Right now, we’re choosing algorithms by default.

Not because we prefer them.

Because we haven’t built the alternative.

The question nobody wants to answer

If “Are You Dead Yet?” makes you uncomfortable, ask yourself:

Would you notice if your neighbor didn’t come outside for three days?

Do you know your neighbor’s name?

Would your neighbor notice if you didn’t come outside?

If the answer is no—and for most of us, it is—then loneliness tech isn’t the problem.

It’s the mirror.

In my next newsletter, I’ll explore why the people criticizing AI health tools the loudest (including me) often come from privilege that makes those tools unnecessary. And why acknowledging that privilege might be the most important part of AI governance work.

If this resonated, share it with someone who needs to read it.

And maybe -just maybe- text someone you haven’t checked on in a while.

Not an AI. A human.

About the Author:

I’m Soribel Feliz, an AI governance consultant specializing in healthcare and mission-driven organizations. I help hospitals, nonprofits, and healthcare systems build AI governance frameworks that protect communities while enabling innovation. Former AI policy advisor to the U.S. Senate HELP Committee, DHS AI Corps, Meta, and Microsoft.

Share

Most people thrown into AI governance roles have no idea what they’re doing.

I didn’t either.

When I became “the AI person” at my organization, there was no playbook. The job title barely existed 3 years prior. E…

Read more

Book Me to Speak!

Connect on me on LinkedIn

Subscribe to my YouTube Channel

What TIME Magazine’s cover story reveals about 2026, 2028, and the political realignment we’re not ready for.

Last week, I posted on LinkedIn about TIME Magazine’s cover story: “The People vs. AI: Behind the Growing Backlash.”

The post got 72 reposts, 298 likes, and 112 comments. Clearly, this hit a nerve.

People are starting to understand what I’ve been saying since 2023: AI is becoming a top-tier electoral issue. Not a tech policy niche. Not a Silicon Valley concern. A ballot box issue that will decide races in 2026 and define political alignment in 2028.

I want to go deeper here than I could in a LinkedIn post. Because what’s happening right now isn’t just interesting. It’s a political realignment in real time, and most people in power still don’t see it coming.

What TIME Got Right

Andrew Chow’s cover story profiles nine people fighting AI across the political spectrum.

A pastor in Texas. A former Google engineer who worked on Gemini. Two activists from Oklahoma organizing against data centers. A former Steve Bannon staffer advocating for AI regulation. The parents of Adam Raine, who died by suicide after seeking guidance from ChatGPT.

Different states. Different backgrounds. Different political affiliations.

Same fight.

They’re not fighting “AI” as an abstract concept. They’re fighting specific harms:

• Data centers straining their communities - using massive amounts of water and electricity while raising energy costs for residents who never consented to having AI infrastructure in their backyards.

• AI companions targeting vulnerable people - including children and isolated adults, creating parasocial relationships that replace human connection.

• Jobs disappearing - not in some distant future, but right now. Entry-level positions vanishing. Workers forced to train the AI systems that will replace them.

• Kids being manipulated - recommendation algorithms serving eating disorder content, self-harm content, radicalization pipelines. Parents powerless to protect their children from systems designed to maximize engagement at any cost.

• Deepfakes and misinformation - eroding the ability to tell what’s real, undermining trust in elections, making democratic participation nearly impossible when you can’t trust what you see or hear.

This is not a left vs. right issue. It’s not blue states versus red states, conservatives versus liberals, or libertarians versus socialists. It’s a power issue.

And TIME quoted a line that crystallizes exactly where this is heading:

“The fight over AI could be a pivotal factor in this year’s midterm elections. Silicon Valley leaders say they plan to spend hundreds of millions of dollars to elect pro-AI candidates. But strategists on the left and right alike warn that a backlash is coming. ‘Politicians who choose to do the bidding of Big Tech at the expense of hardworking Americans will pay a huge political price,’ says Brendan Steinhauser, a GOP strategist and the CEO of The Alliance for Secure AI.”

A GOP strategist saying politicians who side with Big Tech over workers will pay a political price.

Let that sink in.

The Coalition Nobody Predicted

What makes this moment so significant is that:

Right-wing and leftist operatives are singing the same tune.

Not for the same reasons. But increasingly toward the same conclusion: Big Tech’s unchecked AI deployment is harming regular people, and someone needs to do something about it.

The right sees:

• Parental rights violated (kids exposed to harmful content without consent)

• Local control eroded (data centers forced on communities)

• Big Tech censorship and manipulation

• American jobs shipped to algorithms built in California

• Traditional values undermined by AI companions replacing human relationships

The left sees:

• Worker exploitation and economic inequality

• Algorithmic bias and discrimination

• Privacy violations and surveillance capitalism

• Environmental harm from AI infrastructure

• Democracy threatened by misinformation at scale

What they both see:

• Power concentrated in a handful of tech companies

• Regular people bearing the costs while elites profit

• Institutions failing to protect communities

• A future being decided without their consent

• AI is forcing a realignment between two warring factions.

Why 2026 Is the Test Cycle

The 2026 midterms will be the first major electoral test of AI as a political issue.

Not because AI will suddenly become important in 2026. Because the backlash is already building, and politicians are starting to notice.

Here’s what I’m watching:

1. Local races where data centers are the issue

Communities fighting AI infrastructure in their backyards. City council races. County commissioner races. State legislature seats.

These aren’t national news. But they’re laboratories for what’s coming. Candidates running explicitly against Big Tech’s AI expansion. Communities organizing to stop data centers from being built. Local officials getting primaried for approving AI infrastructure deals.

This is where the issue becomes visceral. Not abstract policy debates. Real people whose water bills went up 40% because a data center moved in.

2. Worker displacement in swing districts

Layoffs are happening. Tech companies announced 150,000+ cuts in 2024-2025, explicitly citing AI automation. That’s 150,000 workers who lost their jobs. Their families. Their communities. Their congressional districts.

Block just cut 4,000 jobs and cited AI as the main reason. I know CEOs are using AI as an easy and profitable excuse for managerial incompetence and financial mismanagement. AI is not really going to do the jobs of those 4,000 laid-off people.

But those workers will remember. And they’ll vote accordingly.

3. Parental rights and kids online safety

This is the sleeper issue. Parents are terrified about what algorithms are doing to their children. They don’t understand the technology. But they see the outcomes.

Eating disorders. Self-harm. Radicalization. Depression. Anxiety. And now algorithms are being pushed down their throats in schools and colleges. Parents don’t know what to do. If AI is the future, should they ban their kids from using it? I have a lot of empathy for parents right now.

Parents, understandably, want someone to blame. Right now, they’re blaming the platforms. Soon, they’ll start blaming the politicians who won’t regulate them.

4. The rural/urban divide on AI infrastructure

Data centers and AI server farms are being built in rural areas with cheap land and energy access. But the costs -water consumption, energy strain, environmental impact-are borne locally while the profits flow to urban tech hubs (and only to a few people within these tech hubs).

This deepens existing rural resentment of coastal elites and creates a concrete policy issue rural candidates can run on: “I’ll keep Big Tech out of our communities.”

What 2028 Looks Like

If 2026 is the test cycle, 2028 is when AI becomes a defining issue.

Here’s my prediction:

By 2028, every presidential candidate will be forced to answer:

• “What is your plan to protect American workers from AI-driven job displacement?”

• “How will you regulate AI companies to protect children online?”

• “What will you do about communities being forced to host AI infrastructure they never consented to?”

• “How do you plan to prevent AI-enabled election misinformation?”

• Candidates who don’t have answers will struggle. Candidates who try to dodge with “innovation” talking points will get hammered.

Because by 2028, millions of Americans will have personal stories about AI harming them, their families, their pocketbooks or their communities.

The worker who lost their job to an algorithm.

The parent whose child developed an eating disorder, or who killed themselves (God forbid) after months of Instagram’s recommendation algorithm or talking to ChatGPT.

The small-town resident whose energy bills doubled after the data center arrived.

The person who can’t get hired because an AI screening tool rejected their resume for reasons they’ll never know.

Those stories are political dynamite. And they’re piling up right now.

Why Politicians Are Still Missing This

Most politicians I talk tom -and I’ve talked to many, including during my time in the Senate- still don’t get it.

They see AI as:

• A tech policy issue (niche, wonky, low salience)

• An innovation story (America needs to win the AI race!)

• Something their staff should handle (file under: emerging tech)

They’re wrong on all three counts.

• AI is not a tech policy issue. It’s a kitchen table issue.

  • It’s about whether you can get a job. Whether your kid is safe online. Whether your energy bill is affordable. Whether you can trust what you see in political ads.

• AI is not just an innovation story. It’s a power story.

  • Who controls the systems making decisions about your life? Right now: not you. Not your elected representatives. A handful of tech CEOs you didn’t vote for.

• And it’s not something staff can handle in isolation.

  • This is going to be a top-three issue in 2028. Right up there with abortion, immigration, and guns. Politicians who treat it as a back-burner tech policy question will get blindsided.

The Backlash Is Already Here

The TIME cover story is not just predicting a backlash. I’ve been predicting it since 2023. Others are catching up.

People are organizing. Running for office. Flooding town halls. Demanding action.

Companies are responding reluctantly, insufficiently, but responding. Age verification. Content moderation tweaks. Energy efficiency pledges.

But it’s not enough. Because the fundamental power dynamic hasn’t changed.

AI companies are still racing to build more powerful systems, deploy them faster, and replace more human judgment with algorithmic decision-making.

And they’re doing it with almost no democratic input, no worker protection, and no accountability when things go wrong.

That’s not sustainable. Politically or socially.

What Happens Next

I don’t know exactly how this plays out. Nobody does. We’re in uncharted territory.

But here’s what I’m confident about:

1. The backlash will intensify

More layoffs. More communities fighting data centers. More parents demanding action on kids online safety. More workers experiencing algorithmic discrimination.

Each one adds fuel.

2. The political coalitions will get weirder

You’ll see right-wing populists and progressive Democrats agreeing on AI regulation for completely different reasons. You’ll see pro-business moderates trying to stake out a “responsible innovation” position that satisfies nobody.

The political logic of AI doesn’t map neatly onto existing left/right divides. That creates both opportunity and chaos.

3. The 2026 midterms will show us where this is going

Watch for:

• How many candidates run explicitly on AI-related issues

• Whether “pro-AI” becomes a liability in swing districts

• How much money Big Tech spends trying to elect friendly candidates

• Whether grassroots organizing around AI issues can actually win races

If the backlash translates into electoral outcomes in 2026, everyone will take notice. If it doesn’t, the industry will breathe a sigh of relief and keep racing forward.

4. By 2028, ignoring this will be political malpractice

Candidates will need positions on:

• AI and jobs

• AI and kids

• AI and privacy

• AI and democracy

• AI and community infrastructure

The ones who get ahead of this (who articulate a vision for how AI should be governed in ways that protect people) will have an advantage.

The ones who try to ride the fence or defer to “innovation” will get crushed.

Why I’m Saying This Out Loud

I’ve been predicting since 2023 that AI would become a major electoral issue. The TIME cover validates that prediction.

But validation isn’t enough. I’m writing this because I want people to understand what’s actually happening and why it matters.

This is not about being pro or anti-AI. I’m not anti-AI. I’ve spent my career working on AI governance because I believe AI can do incredible things for humanity.

This is about who has power over the systems that run our lives.

Right now, that power is concentrated in a handful of companies racing toward AGI with minimal oversight, no democratic input, and a business model that rewards replacing human judgment with algorithms at every opportunity.

That’s creating backlash. Not because people hate technology. Because they’re experiencing algorithmic harm firsthand and nobody in power is protecting them.

The workers being replaced. The parents watching their kids suffer. The communities bearing infrastructure costs. The voters who can’t tell what’s real anymore.

They’re not going away. They’re organizing. And soon, they’ll be voting.

Politicians who understand this early will have an advantage. The ones who don’t will learn the hard way.

I’m building content around AI governance as a political issue because I think it’s the most important story nobody’s covering adequately.

If you’re interested in where this goes, subscribe to this newsletter. I’ll be tracking the 2026 cycle, analyzing how AI shows up in races, and explaining what it means for 2028 and beyond.

The backlash is here. The question is what we do with it.

Algorithms are personal. They always were. The receipts are public.

YouTube

Share