RFK Jr.'s AI-Powered Nutrition Website vs. Real Science [2025]

The Contradiction Nobody's Talking About

In January 2025, the U.S. Department of Health and Human Services released new dietary guidelines. Robert F. Kennedy Jr., now leading HHS, promoted them aggressively. The messaging was clear: Americans need more protein, especially from meat. The government even launched a Super Bowl ad featuring Mike Tyson encouraging people to visit a new website called Realfood.gov.

Here's the problem. The website tells people to use Grok, Elon Musk's AI chatbot, for nutrition advice. When you actually ask Grok about the best protein sources, it recommends plant-based proteins, fish, and poultry—and suggests limiting red meat. When asked about RFK Jr.'s carnivore diet, Grok warns about potential scurvy-like symptoms and gout.

So the government's website, promoting the government's protein guidelines, uses an AI that contradicts those exact guidelines.

This isn't a minor inconsistency. It's a window into how AI is being integrated into public health policy, sometimes without a clear understanding of what the AI actually says or whether it aligns with the underlying message. The tension between what Realfood.gov claims and what Grok actually recommends reveals something bigger: we're deploying AI systems to give health advice without really thinking through the consequences.

The stakes matter here. Nutrition advice influences buying decisions, shapes dietary habits, and affects millions of people's health outcomes. When a government health website uses an AI system that contradicts its own recommendations, people lose trust in both the website and the guidelines themselves.

Let's break down what's actually happening, what the science actually says, and why this matters.

The New Protein Guidelines Explained

The administration's new dietary recommendations suggest consuming 1.2 to 1.6 grams of protein per kilogram of body weight daily. For a 70-kilogram person (about 154 pounds), that's 84 to 112 grams per day.

For comparison, the old recommendation was 0.8 grams per kilogram—about 56 grams for that same 154-pound person. So yes, the new recommendation is substantially higher.

The administration frames this as "ending the war on protein." Kennedy has made similar declarations at cattle industry events, stating that "beef is back on the menu." The new food pyramid visual prominently features steak and other animal products at the top.

But here's what the administration's own scientific foundation document says: U.S. adults consume about one gram of protein per kilogram of body weight per day on average. That's already higher than the old 0.8-gram minimum. The same document notes that protein deficiency is "rare" in the United States.

So the new guidelines aren't really responding to a protein deficiency crisis. What's actually happening?

Nutritionists I researched explain that the new guidelines might help people who are metabolically unhealthy—extra protein can increase satiety, helping people feel fuller and eat less overall. Building muscle requires both protein and resistance training, though the messaging sometimes implies that eating more protein alone builds muscle, which isn't how physiology works.

Lindsay Malone, a clinical dietician at Case Western Reserve University, points out the nuance problem: the administration is trying to target people who could benefit from higher protein intake, but the single, simple message gets lost. Then people turn to AI tools like Grok, which provide even more information, potentially overwhelming average users trying to make basic food decisions.

Michelle King Rimer, a clinical assistant professor of nutritional sciences at the University of Wisconsin-Milwaukee, adds another angle: consuming excess protein still gets converted to fat. Too much of any macronutrient leads to weight gain. The message "prioritize protein at every meal" doesn't account for individual caloric needs or activity levels.

The average American consumes about 1.05 g/kg of protein, exceeding the old recommendation of 0.8 g/kg and aligning with the upper range of 1.2 g/kg. Estimated data.

What Most Americans Actually Eat

Let's ground this in actual consumption data. The National Health and Nutrition Examination Survey tracks what Americans actually eat. The data shows that the vast majority of American adults consume adequate protein for basic physiological needs.

Adults 19-50 years old average about 1.0-1.1 grams of protein per kilogram of body weight. That's already in the midpoint of the Acceptable Macronutrient Distribution Range. Adults over 50 consume slightly less but still above the 0.8-gram minimum.

So when the administration says Americans need more protein, they're not describing a widespread deficiency. They're making a different argument: that higher protein intake would be beneficial for specific populations—older adults concerned about muscle loss, people doing strength training, or individuals trying to manage weight through satiety.

That's a reasonable public health position. But it requires nuance. You can't just say "Americans need more protein" without context. You need to say "Americans who want to preserve muscle mass, especially as they age, might benefit from protein intake at the higher end of the recommended range."

The marketing version loses all the nuance.

Grok's Actual Recommendations

When I researched what Grok says about protein sources, the AI consistently recommended a balanced approach: plant-based proteins, fish and seafood, lean poultry, and eggs. It suggested limiting or minimizing red meat and processed meats.

This aligns with major health organizations. The American Heart Association recommends a varied diet with multiple protein sources, emphasizing fish, poultry, legumes, and nuts while limiting red and processed meats. Growing epidemiological evidence shows that diets emphasizing plant-based proteins and fish are associated with better long-term health outcomes than diets high in red meat.

When asked about RFK Jr.'s carnivore diet specifically—meat and fermented foods—Grok mentioned potential benefits like weight loss and reduced brain fog, but also flagged concerns: scurvy-like symptoms (from lack of vitamin C), constipation, and gout.

Those aren't made up. The carnivore diet eliminates whole food groups and can create micronutrient deficiencies. A diet devoid of plant foods removes sources of fiber, which can cause constipation. Purines from excess meat consumption can trigger gout in susceptible individuals. These are legitimate medical concerns, not arbitrary warnings.

So Grok's advice is actually more aligned with mainstream nutritional science than the government's new marketing message. That's bizarre. The government website is telling people to use an AI tool that will give them advice that contradicts what the website itself is promoting.

The new protein guidelines recommend 1.2 to 1.6 grams per kilogram, significantly higher than the old 0.8 grams per kilogram, aligning more closely with current average intake. Estimated data.

The Inconsistency Problem

Michelle King Rimer captures the core issue perfectly: "The inconsistency of the messaging makes it hard for the public to understand what actually matters for their health."

When a government health website tells you one thing, and the AI tool on that website tells you something different, whom do you believe? Most people don't have the nutritional expertise to evaluate competing claims. They'll either trust the government (and ignore Grok), ignore the government's messaging (because Grok said something different), or just become confused and do nothing.

This isn't a theoretical problem. It's a real usability issue. People visiting Realfood.gov looking for guidance leave with conflicting information. They might visit because they saw the Super Bowl ad. They might be trying to improve their health. But the experience won't clarify things—it'll muddy them further.

The administration's response, if asked, might be that Grok just provides information and people should use their own judgment. But that's not how people actually use tools. When you're directed to an AI chatbot on an official government website, you reasonably assume the advice has been vetted.

Why This Happened: The Elon Connection

Why would the HHS website integrate Grok specifically? The answer involves Elon Musk and x AI, the company that created Grok.

Musk has positioned himself as a tech advisor to the administration. He founded x AI in 2023, positioning it as an alternative to Open AI. Grok is the company's flagship AI assistant. In public statements, Musk has claimed Grok is less censored and more willing to engage with controversial topics than other AI systems.

The integration of Grok into Realfood.gov appears to be part of a broader push to promote Musk's AI ventures. That's a business decision disguised as a technical one.

But here's the issue: Grok is a general-purpose AI trained on internet text, including medical literature and health websites. It wasn't fine-tuned specifically for providing government nutrition guidance. It wasn't trained on the new dietary guidelines. It wasn't configured to prioritize the administration's recommendations. It's just... Grok, doing what Grok does.

So when someone asks it about protein sources, it returns its general knowledge, which includes the mainstream scientific consensus that diverse protein sources are healthier than a meat-heavy diet. That consensus exists in its training data because it's supported by decades of research.

The Broader AI in Health Policy Problem

This situation highlights a larger issue: governments are integrating AI into health policy without fully thinking through the implications.

AI systems like Grok are incredibly useful tools. They can synthesize information, explain concepts, and help people learn. But they're not designed for authoritative policy implementation. They make mistakes. They sometimes sound confident while being completely wrong. They reflect the biases in their training data.

Using Grok to "help" people understand nutrition is fine, assuming people understand they're talking to an AI with limitations. But positioning it as the official guidance mechanism on a government health website is different. It implies the AI has been vetted, aligns with policy, and is suitable for that use.

Jessica Knurick, a registered dietitian and public health communicator, has built a social media presence debunking AI-generated nutrition advice. She's clear: "AI gets a lot wrong." She also notes that it's premature to integrate AI guidance into official government health websites without extensive testing and validation.

The FDA has frameworks for medical devices and drug approvals. Those frameworks take years to implement because the stakes are high—people's health is involved. But there's no equivalent framework for AI-powered health guidance on government websites. So we're just... trying it out, in real time, with real citizens.

That's not how evidence-based policy should work.

Younger adults consume more protein compared to older adults, aligning with physiological needs. Estimated data suggests older adults may benefit from increased intake to counter muscle loss.

Protein and Weight Management: The Real Story

Okay, let's zoom back out. Setting aside the AI inconsistency problem, is the new protein recommendation actually good advice?

For some people, yes. For others, it's irrelevant or even counterproductive.

Protein does increase satiety. Eating a high-protein meal makes you feel fuller faster and for longer than eating a high-carbohydrate meal with the same calories. That's well-established. So for people trying to lose weight by reducing overall calorie intake, higher protein consumption can help.

But that's the mechanism. The protein itself doesn't burn fat or build muscle without resistance training. The satiety effect is the benefit. You eat less because you feel fuller, so you're in a caloric deficit, so you lose weight.

Here's the thing, though: you don't need the new 1.2-1.6g/kg recommendation to achieve that effect. Going from 0.8g/kg to 1.0g/kg gets you most of the satiety benefit. Beyond that, you're hitting diminishing returns. The administration's new recommendation goes higher, but not because there's suddenly new evidence that everyone needs that much. It's a broader recommendation targeting multiple populations.

For people who are already eating plenty of protein and aren't trying to lose weight, the new recommendation doesn't change anything. They don't need more protein. They're fine.

For older adults concerned about muscle loss, higher protein intake combined with resistance training is genuinely helpful. That's real evidence-based advice.

For people eating carnivore diets or heavily meat-focused diets, adding even more protein from meat isn't the limiting factor for health—variety is. Eating only one food source, no matter how nutrient-dense, creates deficiencies elsewhere.

The Plant vs. Meat Protein Debate

The administration emphasizes animal protein, particularly beef. The new guidelines state to "prioritize protein at every meal" with "a mix of protein from animal and plant sources." But the rhetoric and visual emphasis, plus Kennedy's cattle industry speeches, make the priority clear.

What does the research actually say about plant versus animal protein?

Both are good sources of amino acids, the building blocks of protein. Animal proteins are "complete proteins," containing all nine essential amino acids in adequate quantities. Most plant proteins are incomplete, missing or low in at least one essential amino acid. But you can combine plant proteins (beans and rice, for example) to get all essential amino acids.

For muscle building and maintenance, both work. Resistance training studies show that people can build muscle on plant protein diets as effectively as on animal protein diets, assuming adequate total protein intake and training stimulus.

Where they differ is in associated health factors. Diets emphasizing plant proteins are associated with:

• Lower cardiovascular disease risk
• Better long-term weight management
• Lower type 2 diabetes risk
• Reduced inflammation markers
• Better gut microbiome diversity

Diets emphasizing red meat are associated with higher cardiovascular disease risk, particularly processed red meat. This isn't controversial—major health organizations including the American Cancer Society and the WHO all recognize this pattern.

So the science says: eat protein from multiple sources, including plants and fish. The administration's new guidelines acknowledge this in the text ("a mix") but emphasize meat in the marketing and imagery.

That's not a scientific disagreement. It's a messaging choice.

Why Grok Got It Right (By Accident)

When Grok recommends diverse protein sources and cautions against excessive red meat, it's not being rebellious. It's just reflecting the scientific consensus in its training data.

The research is consistent and overwhelming. Multiple large prospective cohort studies, systematic reviews, and meta-analyses have reached the same conclusions about protein sources and health. Grok was trained on this body of research. So when it answers a question about protein sources, it outputs what the evidence supports.

The administration could have integrated Grok and also fine-tuned it to prioritize the new guidelines. That's possible. You can use machine learning techniques to adjust an AI's outputs toward specific objectives. But that would mean teaching the AI to recommend more red meat than the evidence supports, potentially for political reasons rather than scientific ones.

Apparently, nobody did that. So Grok just... answers truthfully based on what it learned.

It's almost funny. The government built a platform to promote the administration's dietary guidelines, then filled it with an AI that won't cooperate with the messaging. The AI is more scientifically honest than the marketing is.

Estimated data shows varying protein needs: general population requires around 0.8g/kg, while those in resistance training may need up to 1.7g/kg. Weight loss and older adults also have higher needs.

The Broader Nutrition Advice Crisis

This situation is part of a bigger problem: the American public is flooded with conflicting nutrition information, much of it pushed for economic rather than health reasons.

The beef industry funds research and marketing promoting beef consumption. The dairy industry does the same for dairy. The refined carbohydrate industry has historically done the same (though with less visibility). They're not doing anything illegal—they're just promoting their products.

But when government health policy becomes aligned with industry interests, public trust erodes. People reasonably wonder whose health is actually being prioritized.

The integration of Grok into a government health website, combined with the emphasis on animal protein and the message "beef is back," creates the appearance of industry capture. Whether that's actually the case is debatable. But appearance matters for public health messaging. People need to trust that government health recommendations are based on science, not commerce.

When an AI tool on the government website contradicts the government's marketing, that trust breaks down completely.

How to Actually Evaluate Protein Recommendations

If you're trying to figure out whether you should eat more protein, here's what actually matters:

1. How much protein are you currently eating? Track it for a week. Most people eat more than 0.8g/kg. If you're already above 1.0g/kg, additional protein probably won't change much unless you're trying to lose weight.

2. Are you doing resistance training? If yes, protein at the higher end (1.4-2.0g/kg) supports muscle development. If no, extra protein doesn't build muscle.

3. Are you trying to lose weight? Higher protein increases satiety and helps preserve muscle during caloric deficit. That's legitimate.

4. What's your age and health status? Older adults concerned about muscle loss benefit from higher protein. People with kidney disease need to be careful about protein intake. Context matters.

5. What sources are you eating? Diverse sources—plants, fish, poultry, eggs, and some meat—provide better overall nutrition than focusing on one source.

The administration's new guidelines address some of this. But the marketing simplifies it too much. "Americans need more protein" isn't wrong for everyone, but it's not precisely right for everyone either.

The Limits of AI in Health Communication

Let's address the elephant in the room: should government health websites use AI chatbots for nutrition advice at all?

The arguments for it are straightforward. Chatbots can scale personalized advice. They're available 24/7. They can adjust explanations to different reading levels. They can handle complex follow-up questions.

The arguments against are equally clear:

• AI systems sometimes hallucinate—they generate plausible-sounding information that's completely false
• They reflect the biases in their training data
• They're not transparent about uncertainty (they often sound confident while being wrong)
• They're not designed with the specific context of official policy in mind
• They're not as accountable as human experts are

For basic nutrition information and education, AI can be helpful. But for official health policy guidance, it should be a supplement to human expertise, not a replacement for it.

A better implementation of Realfood.gov would use Grok for exploratory learning—"I want to understand how protein works in my body"—while keeping official policy guidance in human-reviewed, scientifically vetted sections.

As it stands, the website blurs the line. People can't easily tell what's official policy and what's just Grok being Grok.

Jessica Knurick's position is worth taking seriously: "It's premature to be integrating something like this on a government website." We don't have enough evidence yet that AI systems are safe and accurate enough for official health guidance without extensive oversight.

Plant protein diets generally offer better health benefits across various factors compared to meat protein diets. Estimated data based on typical health outcomes.

What Should Actually Change

If the HHS wanted to maintain both the new protein guidelines and use Grok effectively, several things would need to happen:

1. Fine-tune Grok specifically for the new guidelines. Train it on the new recommendations, have it prioritize those messages, and test it extensively.

2. Separate official guidance from exploratory information. Have a clear section for actual government recommendations, with human-reviewed content. Use Grok for answering follow-up questions and exploring concepts, not for delivering policy.

3. Transparency about AI limitations. Tell people they're talking to an AI, explain what it can and can't do, and encourage them to consult registered dietitians for personalized advice.

4. Continuous monitoring and adjustment. If Grok gives bad advice, fix it. Don't just let it run unsupervised.

5. Address the bias in the messaging itself. The "beef is back" framing prioritizes industry interests over science. The guidelines should emphasize evidence-based recommendations without the industry cheerleading.

None of this is impossible. It just requires treating AI integration as a serious matter requiring rigorous oversight, not a cool add-on to a website.

The Bigger Picture: Government, Industry, and Science

This whole situation sits at the intersection of government, industry, and science—three systems that don't always align.

Government has a responsibility to protect public health and communicate evidence-based recommendations. Industry has economic incentives to promote products. Science should be neutral about both.

When government, industry, and science all point in the same direction, great. People get good advice and companies profit from doing the right thing. But when they conflict, government needs to side with science, even if it hurts industry interests.

The new protein guidelines have legitimate scientific support for specific populations. Older adults, people strength training, and people managing weight can all benefit from higher protein intake. That's the science part.

But the marketing and emphasis on beef seems driven by something else—probably the cattle industry's interest in promoting meat consumption. That's the industry part.

And using an AI system that contradicts the government's own messaging? That's the failure part. That's where systems broke down.

What People Should Actually Do

If you're trying to improve your diet and you're confused by conflicting information, here's a practical approach:

Track your current diet for one week. Note protein intake, overall calories, fruits and vegetables consumed, and how you feel energy-wise. Most people are eating enough protein without trying.

Identify your actual goal. Are you trying to lose weight? Build muscle? Have more energy? Feel less sluggish? Different goals need different approaches. "Eat more protein" might help with some but not others.

Focus on the basics first: Eat more vegetables (seriously, most people don't eat nearly enough), drink enough water, move your body regularly, and sleep adequately. These matter far more than fine-tuning protein intake.

Then adjust protein if needed. If you're strength training, eat more protein. If you're trying to lose weight and you're hungry all the time, try increasing protein to improve satiety. If you're older and concerned about muscle loss, aim for the higher end of the range.

Source protein from multiple places. Fish, poultry, eggs, legumes, nuts, seeds, and yes, some meat. Diversity matters.

Don't obsess. Hitting exact macronutrient targets isn't necessary for good health. Being roughly in the right ballpark and focusing on whole foods beats precise tracking with processed foods every time.

If you need personalized advice, see a registered dietitian. That's what they're trained for.

The Long-Term Implications

What happened with Realfood.gov and Grok is likely a preview of how government services will integrate AI over the next decade. Some experiments will work well. Others won't. Grok probably isn't even the most problematic case—there are worse applications of AI to policy domains.

But the basic pattern will repeat: government agencies will integrate AI, sometimes without full consideration of whether the AI aligns with policy, sometimes with insufficient oversight, and sometimes with conflicts of interest mixed in.

That's not an argument against using AI in government. It's an argument for doing it carefully, with proper oversight, scientific validation, and transparency about what the AI can and can't do.

The public health stakes are real. Nutrition advice affects millions of people. When that advice is contradictory or driven by economic interests rather than evidence, people's health suffers.

Realfood.gov could have been a great educational resource. Instead, it's a case study in how not to integrate AI into official health policy.

TL; DR

• New protein guidelines are higher than previous ones, recommending 1.2-1.6g/kg daily versus the old 0.8g/kg minimum, but most Americans already exceed the old minimum
• Grok contradicts the government's messaging, recommending diverse protein sources and cautioning against excess red meat, which aligns with mainstream nutrition science
• The inconsistency confuses public guidance because people don't know whether to follow the government's emphasis on animal protein or the AI's recommendation for diversity
• AI integration needs serious oversight when used for official health policy, including fine-tuning, testing, monitoring, and transparency about limitations
• The research supports varied protein sources, particularly emphasizing plant proteins and fish while limiting red meat, yet the government's marketing emphasizes beef

FAQ

How much protein does the average American actually eat?

Research shows that most American adults consume approximately 1.0 to 1.1 grams of protein per kilogram of body weight daily, which already exceeds the old 0.8g/kg minimum recommendation and meets the midpoint of the Acceptable Macronutrient Distribution Range. This means protein deficiency is rare in the United States, though specific populations like older adults might benefit from protein intake at the higher end of recommended ranges.

What does Grok actually recommend for protein sources?

When asked about protein sources, Grok recommends a balanced approach including plant-based proteins, fish and seafood, lean poultry, and eggs, while suggesting people limit or minimize red meat and processed meats. This recommendation aligns with guidance from major health organizations like the American Heart Association and reflects the scientific consensus based on epidemiological research.

Can you build muscle on a plant-based protein diet?

Yes, absolutely. Research on resistance training shows that people can build muscle effectively on plant-based protein diets as long as they consume adequate total protein and engage in appropriate strength training. While animal proteins are complete proteins containing all essential amino acids, plant proteins can be combined (such as beans with rice) to provide all necessary amino acids for muscle development.

Why did the government integrate Grok into Realfood.gov if it contradicts the messaging?

The integration appears driven by the administration's alignment with Elon Musk and x AI rather than by careful evaluation of whether Grok's recommendations match the government's guidelines. The system wasn't fine-tuned specifically for the new dietary recommendations, so Grok simply returns its general knowledge, which reflects mainstream nutritional science rather than the government's marketing emphasis on animal protein.

Is the new protein recommendation actually beneficial for everyone?

Not necessarily. The higher protein recommendation is particularly beneficial for older adults concerned about age-related muscle loss, people engaged in regular resistance training, and individuals trying to lose weight through satiety-driven calorie reduction. For people already consuming adequate protein who aren't strength training, additional protein provides minimal additional benefit.

What's the actual relationship between protein and weight loss?

Protein increases satiety, making you feel fuller faster and for longer than equivalent calories from carbohydrates or fats. This helps with weight loss because you eat fewer total calories, not because protein itself burns fat or has special metabolic properties. You still need to be in a caloric deficit for weight loss to occur.

Should I trust AI chatbots for nutrition advice?

AI chatbots can provide general education and explanations about nutrition, but they shouldn't be your primary source for personalized nutrition advice, especially regarding health conditions or specific medical situations. Registered dietitians have training and credentialing that AI systems lack, and they can account for your individual health status, medications, and goals in ways general AI cannot.

How can I tell if nutrition advice is driven by industry interests versus science?

Consider who's promoting the advice and what they stand to gain. Advice promoting products from the organization giving it is automatically suspect, even if it happens to be scientifically accurate. The most trustworthy advice comes from sources without financial interest in specific products, such as academic researchers, registered dietitians without corporate ties, or government health agencies (when they're not captured by industry).

What does the research actually say about red meat consumption?

Multiple large prospective studies and meta-analyses show that high consumption of red meat, especially processed red meat, is associated with increased cardiovascular disease risk, type 2 diabetes, and certain cancers. Replacing just 10% of red meat calories with plant protein, fish, or poultry is associated with significant risk reduction, according to Harvard Medical School researchers.

Is RFK Jr.'s carnivore diet actually healthy?

A carnivore diet eliminates entire nutrient categories, creating significant risk of deficiencies in vitamin C (increasing scurvy risk), fiber (causing digestive issues), and various micronutrients found primarily in plant foods. While some people report short-term weight loss and improved energy, the long-term health impacts are not well-studied, and the approach contradicts established nutritional science regarding dietary variety.

Final Thoughts

The Realfood.gov situation matters because it reveals how government health policy intersects with AI integration, industry interests, and scientific integrity. The contradiction between the website's messaging and Grok's recommendations isn't just a glitch—it's a symptom of decisions made without careful consideration of the consequences.

Government health websites need to be trustworthy. People visiting them should encounter consistent, evidence-based guidance. When that guidance conflicts with itself, people rightfully lose confidence in the entire system.

The solution isn't to ban AI from government services. It's to integrate AI carefully, with proper oversight, scientific validation, and transparency about what the AI can and can't do. Until we do that, we'll keep seeing these contradictions, and public health will suffer as a result.

Nutrition science is complex enough without adding conflicting AI recommendations into the mix. People deserve clarity.

Key Takeaways

• Realfood.gov uses Grok to provide nutrition guidance, but the AI recommends diverse protein sources while the government emphasizes animal protein, creating direct contradictions
• Most Americans already consume adequate protein (averaging 1.0-1.1g/kg daily) exceeding the old 0.8g/kg minimum, so the new guidelines target specific populations rather than addressing widespread deficiency
• Grok's recommendations for plant-based proteins and cautions against excessive red meat align with mainstream nutritional science and health organizations, contradicting the government's marketing emphasis
• AI systems integrated into official government health policy require fine-tuning, extensive testing, and continuous monitoring to ensure alignment with stated guidelines—Realfood.gov doesn't appear to have done this
• The situation reveals potential industry capture in health policy, with the cattle industry's interests appearing to influence messaging beyond what scientific evidence supports

Related Articles

• RFK Jr.'s Autism Panel & Vaccine Conspiracy Theories [2025]
• The Fake War on Protein: Politics, Masculinity & Nutrition [2025]