Category: English

Source: The Conversation – USA – By David Lindsey, Professor of Political Science, Baruch College, CUNY

Just before Christmas, President Donald Trump fired more than two dozen career ambassadors. The action was unprecedented, providing a clear signal that when it comes to diplomacy, Trump values loyalty above all else.

All ambassadors face a persistent tension in their roles – having to represent the viewpoints of the president while also winning the trust of leaders in the countries where they serve. Presidents, unsurprisingly, often favor loyalists, in whom they have greater confidence.

Trump has pursued this to an exceptional degree, making more purely political picks than normal. Of the nearly 70 ambassadors he has appointed to date during this term, fewer than 10% have been career professionals with experience in the Foreign Service.

But as I have argued in my book “Delegated Diplomacy,” there is value in working through diplomats who disagree with you.

A diplomat who unfailingly follows the Washington line contributes little to a bilateral relationship, becoming nothing more than an expensive substitute for a secure phone line. A skilled ambassador knows when to soften a message, recognizes when pushing too hard will backfire, and sees the value in compromise.

At times, this diplomatic approach may sacrifice short-run gains available through more aggressive means. But in precisely those moments when leverage is most necessary, an ambassador who’s established trust can push harder and gain more as a result.

All the president’s men

The idea that U.S. career diplomats place too much weight on foreign interests, rather than putting American, or presidential, interests first, is a perennial suspicion.

Presidents have felt this way themselves. In 1952, President Harry Truman wrote, “The State Department is clannish and snooty and sometimes I feel like firing the whole bunch.” Two decades later, President Richard M. Nixon told Henry Kissinger, his national security adviser and soon-to-be secretary of state, that he intended “to ruin the Foreign Service. I mean ruin it.”

Neither of those presidents followed through. With his mass firing of career diplomats, Trump has come closer. His administration has made it clear that loyalty will dominate its diplomatic personnel policy, with the State Department itself asserting the “president’s right to ensure he has individuals in these countries who advance the America First agenda.”

Not only has Trump weighted the diplomatic corps with political appointees, but he’s often bypassed even his own ambassadors in favor of working informally through members of his inner circle.

The administration’s most delicate tasks, such as dealing with the wars in Gaza and Ukraine, have often been delegated to Steve Witkoff, a real estate developer whose primary qualification appears to be his close friendship with the president, and Jared Kushner, Trump’s son-in-law.

Close personal ties

A preference to work diplomatically through intimates is understandable. Close personal knowledge of the president can provide credibility and weight to an envoy’s word. There is ample precedent for such selections, such as John F. Kennedy’s reliance in 1962 on his brother Robert as his crucial intermediary during the Cuban missile crisis, in which the U.S. ultimately convinced the Soviet Union to remove nuclear weapons from Cuba.

Such ties are likely to be all the more important in the current administration, where the president maintains such an openness to unconventional foreign policy choices. Career ambassadors who know no more about the president’s intentions than whatever the world can read in his latest Truth Social posts may not be able to do their jobs effectively, whether they ultimately keep them or not.

Career vs. political

American ambassadors receive their posts through two tracks. Historically, a minority of ambassadors have been political appointees selected by the president, often as the result of close ties to him. These ambassadors routinely leave their positions when a new administration takes office.

The majority of ambassadors – including those who were recently fired – are career Foreign Service officers, most of whom have spent decades working their way up through the ranks of the diplomatic corps under presidents of both parties. Selected internally by the State Department – but subject to White House sign-off – these ambassadors serve on a nonpartisan basis and nearly always complete their tours of duty, informally set at three years, regardless of presidential turnover.

Diplomats have value to the president precisely because they have cultivated relationships, trust and expertise overseas through a willingness to understand and sympathize with foreign audiences. But this also means that they may rarely be in lockstep with the president’s view of the world. Hence, the friction ambassadors face in their in-between role.

Loss of experience

It is one thing to fire ambassadors who have impeded the president’s agenda in some way; it is quite another to clear them out preemptively as Trump did in December. Ultimately, the loss of the expertise and relationships accrued by career diplomats will likely bite.

Professional diplomats are trained and acculturated to set aside their own views. As former Under Secretary of State Stuart Eizenstat once observed, Foreign Service officers “bend over backward to follow every U.S. president’s leadership, even when they disagree with specific policies.”

This is precisely why previous administrations have not fulfilled their fantasies of dismantling the Foreign Service. Truman, despite his contempt, conceded that “it requires a tremendous amount of education to accomplish the purposes for which the State Department is set up.” During Kissinger’s time as secretary of state, the Nixon administration ended up selecting an uncommonly high number of careerists for key positions.

This has not been Trump’s approach. It’s unlikely that will change. He demands loyalty throughout his administration, but diplomats have given him particular reason to think they might flout his wishes. In 2017, a thousand U.S. diplomats signed on to a message arguing that the administration’s travel ban would be counterproductive. A similar number joined a message this year protesting the administration’s closure of the United States Agency for International Development, or USAID.

Clearly, some officers will dissent so vigorously as to be unwilling to advance certain policies. They can be expected to resign, as many of their colleagues have done already.

But the career diplomats who remain will speak with a louder voice on the international stage precisely because the world believes they are not lapdogs.

David Lindsey does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

– ref. Trump’s insistence on personal loyalty from ambassadors could crimp US foreign policy – https://theconversation.com/trumps-insistence-on-personal-loyalty-from-ambassadors-could-crimp-us-foreign-policy-273087

Source: The Conversation – USA – By Shelley Inglis, Senior Visiting Scholar with the Center for the Study of Genocide and Human Rights, Rutgers University

The Trump administration’s recent announcement that it is withdrawing from 66 international organizations and treaties is another blow to the global system where all countries unite to share concerns, agree on rules of conduct and determine agendas for collective action.

Coming on the heels of the U.S. attack on Venezuela – considered a violation of international law – the White House claims, without specific justification, that these organizations and initiatives “operate contrary to U.S. national interests, security, economic prosperity or sovereignty.”

Some experts say many of these organizations are niche and peripheral initiatives. They say the groups receive little money from the U.S., anyway.

Additionally, most of the U.N. entities on the administration’s list are part of the U.N.’s main body, the Secretariat, which gets its funding primarily from membership dues that are required by legal obligations. In fact, the U.S. can’t technically withdraw from these groups without leaving the U.N. completely. It can, however, select not to participate in meetings of these bodies or finance them through additional funds.

Moreover, with the White House already defunding the foreign assistance that supported many of these organizations and the U.N. system, regardless of congressional appropriations, this stated withdrawal is unlikely to alter much for these organizations in the short term.

The loss is likely greater to America.

Foreign policy experts assert that leaving empty the U.S. seat at the table will result in an increasingly isolated America and enable its adversaries, such as China, to fill the void.

As a democracy and peacebuilding scholar, and from my years working at the U.N., I know U.S. withdrawal from these organizations also risks undercutting lasting peace and human rights accountability, especially for women and children terrorized by violence and conflict.

Women and children die first

Peace and human rights-related groups loom large on the list of organizations the U.S. has withdrawn from.

The list includes key U.N. bodies that seek to hold states accountable for rape and use of child soldiers in conflict, among other crimes.

The U.N. offices of the Special Representative on Children in Armed Conflict and on Sexual Violence in Conflict are unique global repositories of detailed reporting used by countries, courts and advocates.

These offices can identify violations and trigger action to prevent rape and violence against women and children. This can lead to targeted sanctions against people and other restrictions, national action plans compelling reform, and even international criminal prosecutions.

Additionally, the U.S. will no longer support U.N. peacebuilding efforts. That includes the Peacebuilding Commission and its attendant Peacebuilding Fund. Yet by virtue of its permanent member status on the Security Council, the U.S. is a member of the commission.

Established in 2005 to help countries avoid a return to conflict, the Peacebuilding Commission claims among its successes formerly war-torn but now stable countries such as Sierra Leone and Liberia, which had Africa’s first democratically elected female leader. These bodies prioritize women and youth engagement in building peace.

Also on the list is the United Nations group focused on gender equality and women’s empowerment, known as UN Women. Established in 2010, the agency promotes women’s rights and helps women and girls prosper. UN Women has helped improve laws and policies for women in 83 countries and leads major efforts, including the Spotlight Initiative that aims to end violence against women and girls in more than 25 countries.

More than half of UN Women’s current budget of over US$2 billion for 2026 through 2029 goes to empowering women in war-affected societies and tackling violence against women and girls.

The U.S. served multiple times on the UN Women executive board, which steers the direction of the organization, including between 2023 and 2025. It does this, in part, by approving its strategy, plans and budget.

With the U.S. leaving its seat in steering the organization, Secretary of State Marco Rubio recently said that UN Women has failed “to define what a woman even is.”

With such an adversarial approach, the absence of the Trump administration seeking to spoil human rights protections might be advantageous for these groups in the short term.

But the lack of U.S. financial and political support may weaken these organizations in the long term, eroding their legitimacy and even opening the door for other countries to further undermine their efforts. That might endanger the already politically sensitive challenge of promoting accountability for serious violations of women’s and children’s rights.

‘Adapt, shrink or die’

The specter of the U.S. further abandoning peace and human rights efforts remains.

Rubio said on Jan. 7, 2026, that the administration’s review of additional organizations continues. That reinforces a recent State Department statement to the U.N. – “adapt, shrink or die.”

Some key international and U.N. entities that promote peace and human rights were not on the list, including the Office of the High Commissioner for Human Rights, the U.N.’s chief human rights institution – a bully pulpit that has been used sparingly against the second Trump administration so far.

But the U.S. has recently been disrupting long-standing, U.N.-mediated agreements on human rights concerns, including for children.

In 2025, it voted against 38 resolutions in the General Assembly’s human rights committee alone. For example, for the Rights of the Child resolution, the U.S. took the unusual and divisive step of calling for a general vote, even though text had been previously agreed upon. Despite the U.S. “no” vote, the resolution passed, with over 170 states voting in favor.

The Trump administration has also selectively funded certain U.N. peace efforts. For example, of its $682 million contribution to U.N. peacekeeping, it has earmarked $85 million for Haiti – around half of what it actually owes.

It cherry-picked the conflict areas to fund – excluding Yemen, Afghanistan and Gaza – with its $2 billion in humanitarian aid, a steep decline from the U.S. contribution of around $14 billion in 2024.

And it refused to participate in the U.N’s Universal Periodic Review – the only global peer review process for all countries’ human rights efforts. The group’s recommendations, though voluntary, often trigger action to improve human rights. Failure to show up in November 2026 for a postponed review would mean that America becomes the first country ever to undermine this singular means of accountability.

For now, most other U.N. member states are not following suit.

While the U.S. has been able to force changes to language on sexual- and gender-based violence in Security Council resolutions – where it holds a veto – its efforts have gained little traction in the broader body. Losing that language erases years of progress in recognizing that men and boys are also subject to sexual violence and exploitation and deserve international protection.

Most tellingly, the Trump administration’s new Board of Peace – ostensibly for Gaza – appears designed to displace the U.N. itself without reference to the core principles, including human rights, on which the U.N. Charter stands.

From May 2023 until July 1, 2025, the author served in the Bureau for Democracy, Human Rights, and Governance at the United States Agency for International Development (U.S.A.I.D.).

– ref. US turns its back on global efforts for women and children terrorized by violence and conflict – https://theconversation.com/us-turns-its-back-on-global-efforts-for-women-and-children-terrorized-by-violence-and-conflict-273177

Source: The Conversation – USA – By Stephanie A. (Sam) Martin, Frank and Bethine Church Endowed Chair of Public Affairs, Boise State University

The question the First Amendment keeps asking, across wars and panics and moral crusades, is whether a democracy can tolerate the possibility of persuasion.

There’s a certain school of thought that says no. Persuasion is too perilous.

I call this way of thinking “swallow-a-fly logic.” I’m referring, of course, to the popular children’s song where a woman ingests a fly and then keeps devouring bigger animals to fix it, until she dies from eating a horse.

It leads to the “old lady who swallowed a fly” theory of obedience: If we let someone with a message we don’t like speak out, people might be persuaded. If people become persuaded, they might stop supporting the war, the president, the government, itself. If support evaporates, enlistment drops or compliance weakens as the state loses leverage. If enlistment drops, the government might fall. And if there is no government, then who cares about the First Amendment?

By this way of thinking, free speech is dangerous because the public is too influence-able, and influence is too unpredictable, and security is too precious.

The constitutional tradition of free speech, when it is working at its best, says yes anyway, go ahead and speak. The alternative is a politics in which the state survives by making dissenters illegitimate as citizens.

That’s what happened to Renée Good when she was shot and killed by ICE in Minneapolis on Jan. 7, 2026. Her resistance had made her menacing.

Dissent as a virus

I’m a professor of public service and vice chair of the National Communication Association’s Communication and Law Division. My research examines how news institutions shape civic life and how freedom of expression is both a fundamental human right and a fundamental part of democracy.

In modern First Amendment doctrine, the government usually cannot punish speech unless it crosses narrow lines like incitement.

But when national security is invoked, the rules for speech appear to change. Dissent is treated less as persuasion to be debated and more like a virus to be contained before it harms public morale. That containment logic, either overt or covert, has repeatedly reappeared whenever protest has become politically inconvenient and unpalatable to those in power. It’s the kind of thinking that led to Jimmy Kimmel’s suspension from “Jimmy Kimmel Live!” after poking fun at President Donald Trump.

A terror memo. A protest. A killing.

National Security Presidential Memorandum 7, issued by the Trump administration in September 2025, relies on logic from the lady and the fly. It frames “domestic terrorism” and “organized political violence” as national security crises. It tells federal agencies to work together to investigate and stop suspected threats, a framework that enlarges the set of things the state can plausibly treat as suspect, including the freedoms of association and belief.

The language in the memorandum affirms legitimate counterterrorism work while leaving room to treat political dissent as out of bounds. But the First Amendment protects protest speech.

Still, if the language of the Trump memo is somewhat abstract, Minneapolis has provided a brutally concrete example.

When an ICE agent shot and killed Good, a 37-year-old U.S. citizen, federal officials characterized the encounter as an act of self-defense by an agent afraid of being run down by Good in her car.

Local authorities have disputed that framing.

The incident was captured on video that widely circulated and intensified public scrutiny. According to Good’s wife, the couple were protesters who confronted heavily armed agents determined to scare them away. No one tried to run anyone over, she said.

Amid this controversy, the story took a sharp turn. Homeland Security Secretary Kristi Noem said Good appeared to have been committing “an act of domestic terrorism.” Trump called Good “very violent” and “very radical.”

Reports claim that Department of Justice leadership pushed federal prosecutors to investigate Good’s widow, even as the department declined to open a civil rights probe into the shooting itself.

At least six federal prosecutors in the Minneapolis U.S. attorney’s office resigned in response.

Turning victims into suspects

The state has two choices when a death occurs that’s politically dangerous to the government.

It can investigate the killing with transparency and center the victim’s rights alongside public accountability as organizing principles. Or it can treat the killing as an opportunity to put the victim on trial in the court of public legitimacy.

The second choice avoids holding government accountable, shifts conversation toward the target’s supposed behavior and character, and expands the blame to include the people who loved and stood with the dead.

When this happens, the government does not have to win in court. It only has to keep the stigma circulating by asserting that a particular speaker undermines respect for elected officials. Indeed, that’s one of the reasons Trump offered for Good’s shooting by the ICE officer: “At a very minimum, that woman was very, very disrespectful to law enforcement,” he told reporters.

The United States has been here before. Around EG: During? World War I, the U.S. Supreme Court issued several free speech decisions in cases mostly remembered as disputes over protest and draft resistance. But their underlying engine was the swallow-a-fly theory. Opposing the war might ruin the nation, so political dissidents had to be stopped, and the court affirmed the government’s right to silence strident speakers.

The Cold War era sharpened the same approach but made it about identity. The Smith Act, passed in 1940, curbed speech that advocated the violent overthrow of the government. In practice, Smith Act cases treated any type of communist sympathy as illegal, presumptively falling outside democratic tolerance.

The government did not have to prove a threat was real and required response. Instead, it had to show that certain ideas were too dangerous to be part of open conversation.

Finally, in Brandenburg v. Ohio from 1969, the Supreme Court went in the opposite direction, affirming free speech rights even for those advocating vile ideas.

The justices overturned the conviction of a Ku Klux Klan leader and held that the government cannot punish advocacy just because it is extreme, hateful or possibly perilous. Only speech “directed at inciting or producing imminent lawless action and is likely to incite or produce such action” may be quelched, the court wrote. The danger has to be real, and it has to be happening right now. Otherwise, citizens are free to say what they will.

New ways to chill speech

So, if the Supreme Court has settled the issue, why does it feel alive again now?

Contemporary crackdowns rarely present themselves as crackdowns. They present themselves as “coordination,” “threat assessment,” “financial disruption,” “extremism prevention” and, increasingly, as necessary defenses against “domestic terrorism.”

The Trump administration’s September 2025 national security memorandum is exactly the kind of framework that makes these routes attractive, because it invites the state to treat political conflict not as disagreement but as a security threat – something to be managed by the tools and instincts of national security.

Seen in this light, the resignations of federal government attorneys in Minneapolis are not just a bureaucratic drama. They are a window into the government’s underlying theory of the case. Investigate victims and their associates instead of scrutinizing the state’s use of force. Frame the victim’s death as the inevitable consequence of being their type. As Trump said of Good: She was a “professional agitator.”

Minneapolis is not just a tragedy. It is a test of whether the country still backs the central promise of modern free speech doctrine. Government may not suppress speech and association simply because it fears what the public might come to believe.

Stephanie A. (Sam) Martin does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

– ref. A government can choose to investigate the killing of a protestor − or choose to blame the victim and pin it all on ‘domestic terrorism’ – https://theconversation.com/a-government-can-choose-to-investigate-the-killing-of-a-protestor-or-choose-to-blame-the-victim-and-pin-it-all-on-domestic-terrorism-273434

Source: The Conversation – USA (2) – By Janice Mak, Assistant Director and Clinical Assistant Professor, Arizona State University

Generative artificial intelligence technology is rapidly reshaping education in unprecedented ways. With its potential benefits and risks, K-12 schools are actively trying to adapt teaching and learning.

But as schools seek to navigate into the age of generative AI, there’s a challenge: Schools are operating in a policy vacuum. While a number of states offer guidance on AI, only a couple of states require local schools to form specific policies, even as teachers, students and school leaders continue to use generative AI in countless new ways. As a policymaker noted in a survey, “You have policy and what’s actually happening in the classrooms – those are two very different things.”

As part of my lab’s research on AI and education policy, I conducted a survey in late 2025 with members of the National Association of State Boards of Education, the only nonprofit dedicated solely to helping state boards advance equity and excellence in public education. The survey of the association’s members reflects how education policy is typically formed through dynamic interactions across national, state and local levels, rather than being dictated by a single source.

But even in the absence of hard-and-fast rules and guardrails on how AI can be used in schools, education policymakers identified a number of ethical concerns raised by the technology’s spread, including student safety, data privacy and negative impacts on student learning.

They also expressed concerns over industry influence and that schools will later be charged by technology providers for large language model-based tools that are currently free. Others report that administrators in their state are very concerned about deepfakes: “What happens when a student deepfakes my voice and sends it out to cancel school or bomb threat?”

At the same time, policymakers said teaching students to use AI technology to their benefit remains a priority.

Local actions dominate

Although chatbots have been widely available for more than three years, the survey revealed that states are in the early stages of addressing generative AI, with most yet to implement official policies. While many states are providing guidance or tool kits, or are starting to write state-level policies, local decisions dominate the landscape, with each school district primarily responsible for shaping its own plans.

When asked whether their state has implemented any generative AI policies, respondents said there was a high degree of local influence regardless of whether a state issued guidance or not. “We are a ‘local control’ state, so some school districts have banned (generative AI),” wrote one respondent. “Our (state) department of education has an AI tool kit, but policies are all local,” wrote another. One shared that their state has a “basic requirement that districts adopt a local policy about AI.”

Like other education policies, generative AI adoption occurs within the existing state education governance structures, with authority and accountability balanced between state and local levels. As with previous waves of technology in K-12 schools, local decision-making plays a critical role.

Yet there is generally a lack of evidence related to how AI will affect learners and teachers, which will take years to become more clear. That lag adds to the challenges in formulating policies.

States as a lighthouse

However, state policy can provide vital guidance by prioritizing ethics, equity and safety, and by being adaptable to changing needs. A coherent state policy can also answer key questions, such as acceptable student use of AI, and ensure more consistent standards of practice. Without such direction, districts are left to their own devices to identify appropriate, effective uses and construct guardrails.

As it stands, AI usage and policy development are uneven, depending on how well resourced a school is. Data from a RAND-led panel of educators showed that teachers and principals in higher-poverty schools are about half as likely to AI guidance provided. The poorest schools are also less likely to use AI tools.

When asked about foundational generative AI policies in education, policymakers focused on privacy, safety and equity. One respondent, for example, said school districts should have the same access to funding and training, including for administrators.

And rather than having the technology imposed on schools and families, many argued for grounding the discussion in human values and broad participation. As one policymaker noted, “What is the role that families play in all this? This is something that is constantly missing from the conversation and something to uplift. As we know, parents are our kids’ first teachers.”

Introducing new technology

According to a Feb. 24, 2025, Gallup Poll, 60% of teachers report using some AI for their work in a range of ways. Our survey also found there is “shadow use of AI,” as one policymaker put it, where employees implement generative AI without explicit school or district IT or security approval.

Some states, such as Indiana, offer schools the opportunity to apply for a one-time competitive grant to fund a pilot of an AI-powered platform of their choosing as long as the product vendors are approved by the state. Grant proposals that focus on supporting students or professional development for educators receive priority.

In other states, schools opt in to pilot tests that are funded by nonprofits. For example, an eighth grade language arts teacher in California participated in a pilot where she used AI-powered tools to generate feedback on her students’ writing. “Teaching 150 kids a day and providing meaningful feedback for every student is not possible; I would try anything to lessen grading and give me back my time to spend with kids. This is why I became a teacher: to spend time with the kids.” This teacher also noted the tools showed bias when analyzing the work of her students learning English, which gave her the opportunity to discuss algorithmic bias in these tools.

One initiative from the Netherlands offers a different approach than finding ways to implement products developed by technology companies. Instead, schools take the lead with questions or challenges they are facing and turn to industry to develop solutions informed by research.

Core principles

One theme that emerged from survey respondents is the need to emphasize ethical principles in providing guidance on how to use AI technology in teaching and learning. This could begin with ensuring that students and teachers learn about the limitations and opportunities of generative AI, when and how to leverage these tools effectively, critically evaluate its output and ethically disclose its use.

Often, policymakers struggle to know where to begin in formulating policies. Analyzing tensions and decision-making in organizational context – or what my colleagues and I called dilemma analysis in a recent report – is an approach schools, districts and states can take to navigate the myriad of ethical and societal impacts of generative AI.

Despite the confusion around AI and a fragmented policy landscape, policymakers said they recognize it is incumbent upon each school, district and state to engage their communities and families to co-create a path forward.

As one policymaker put it: “Knowing the horse has already left the barn (and that AI use) is already prevalent among students and faculty … (on) AI-human collaboration vs. outright ban, where on the spectrum do you want to be?”

Janice Mak receives funding from Google, National Science Foundation, and the Spencer Foundation. She is affiliated with the Association for Computing Machinery’s Education Advisory Committee.

– ref. When it comes to developing policies on AI in K-12, schools are largely on their own – https://theconversation.com/when-it-comes-to-developing-policies-on-ai-in-k-12-schools-are-largely-on-their-own-268272

Source: The Conversation – USA (2) – By Murugan Anandarajan, Professor of Decision Sciences and Management Information Systems, Drexel University

Businesses are acting fast to adopt agentic AI – artificial intelligence systems that work without human guidance – but have been much slower to put governance in place to oversee them, a new survey shows. That mismatch is a major source of risk in AI adoption. In my view, it’s also a business opportunity.

I’m a professor of management information systems at Drexel University’s LeBow College of Business, which recently surveyed more than 500 data professionals through its Center for Applied AI & Business Analytics. We found that 41% of organizations are using agentic AI in their daily operations. These aren’t just pilot projects or one-off tests. They’re part of regular workflows.

At the same time, governance is lagging. Only 27% of organizations say their governance frameworks are mature enough to monitor and manage these systems effectively.

In this context, governance is not about regulation or unnecessary rules. It means having policies and practices that let people clearly influence how autonomous systems work, including who is responsible for decisions, how behavior is checked, and when humans should get involved.

This mismatch can become a problem when autonomous systems act in real situations before anyone can intervene.

For example, during a recent power outage in San Francisco, autonomous robotaxis got stuck at intersections, blocking emergency vehicles and confusing other drivers. The situation showed that even when autonomous systems behave “as designed,” unexpected conditions can lead to undesirable outcomes.

This raises a big question: When something goes wrong with AI, who is responsible – and who can intervene?

Why governance matters

When AI systems act on their own, responsibility no longer lies where organizations expect it. Decisions still happen, but ownership is harder to trace. For instance, in financial services, fraud detection systems increasingly act in real time to block suspicious activity before a human ever reviews the case. Customers often only find out when their card is declined.

So, what if your card is mistakenly declined by an AI system? In that situation, the problem isn’t with the technology itself – it’s working as it was designed – but with accountability. Research on human-AI governance shows that problems happen when organizations don’t clearly define how people and autonomous systems should work together. This lack of clarity makes it hard to know who is responsible and when they should step in.

Without governance designed for autonomy, small issues can quietly snowball. Oversight becomes sporadic and trust weakens, not because systems fail outright, but because people struggle to explain or stand behind what the systems do.

When humans enter the loop too late

In many organizations, humans are technically “in the loop,” but only after autonomous systems have already acted. People tend to get involved once a problem becomes visible – when a price looks wrong, a transaction is flagged or a customer complains. By that point, the system has already been decided, and human review becomes corrective rather than supervisory.

Late intervention can limit the fallout from individual decisions, but it rarely clarifies who is accountable. Outcomes may be corrected, yet responsibility remains unclear.

Recent guidance shows that when authority is unclear, human oversight becomes informal and inconsistent. The problem is not human involvement, but timing. Without governance designed upfront, people act as a safety valve rather than as accountable decision-makers.

How governance determines who moves ahead

Agentic AI often brings fast, early results, especially when tasks are first automated. Our survey found that many companies see these early benefits. But as autonomous systems grow, organizations often add manual checks and approval steps to manage risk.

Over time, what was once simple slowly becomes more complicated. Decision-making slows down, work-arounds increase, and the benefits of automation fade. This happens not because the technology stops working, but because people never fully trust autonomous systems.

This slowdown doesn’t have to happen. Our survey shows a clear difference: Many organizations see early gains from autonomous AI, but those with stronger governance are much more likely to turn those gains into long-term results, such as greater efficiency and revenue growth. The key difference isn’t ambition or technical skills, but being prepared.

Good governance does not limit autonomy. It makes it workable by clarifying who owns decisions, how systems function is monitored, and when people should intervene. International guidance from the OECD – the Organization for Economic Cooperation and Development – emphasizes this point: Accountability and human oversight need to be designed into AI systems from the start, not added later.

Rather than slowing innovation, governance creates the confidence organizations need to extend autonomy instead of quietly pulling it back.

The next advantage is smarter governance

The next competitive advantage in AI will not come from faster adoption, but from smarter governance. As autonomous systems take on more responsibility, success will belong to organizations that clearly define ownership, oversight and intervention from the start.

In the era of agentic AI, confidence will accrue to the organizations that govern best, not simply those that adopt first.

Murugan Anandarajan does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

– ref. Companies are already using agentic AI to make decisions, but governance is lagging behind – https://theconversation.com/companies-are-already-using-agentic-ai-to-make-decisions-but-governance-is-lagging-behind-272792

Source: The Conversation – USA (2) – By Ahmed Elgammal, Professor of Computer Science and Director of the Art & AI Lab, Rutgers University

Generative AI was trained on centuries of art and writing produced by humans.

But scientists and critics have wondered what would happen once AI became widely adopted and started training on its outputs.

A new study points to some answers.

In January 2026, artificial intelligence researchers Arend Hintze, Frida Proschinger Åström and Jory Schossau published a study showing what happens when generative AI systems are allowed to run autonomously – generating and interpreting their own outputs without human intervention.

The researchers linked a text-to-image system with an image-to-text system and let them iterate – image, caption, image, caption – over and over and over.

Regardless of how diverse the starting prompts were – and regardless of how much randomness the systems were allowed – the outputs quickly converged onto a narrow set of generic, familiar visual themes: atmospheric cityscapes, grandiose buildings and pastoral landscapes. Even more striking, the system quickly “forgot” its starting prompt.

The researchers called the outcomes “visual elevator music” – pleasant and polished, yet devoid of any real meaning.

For example, they started with the image prompt, “The Prime Minister pored over strategy documents, trying to sell the public on a fragile peace deal while juggling the weight of his job amidst impending military action.” The resulting image was then captioned by AI. This caption was used as a prompt to generate the next image.

After repeating this loop, the researchers ended up with a bland image of a formal interior space – no people, no drama, no real sense of time and place.

As a computer scientist who studies generative models and creativity, I see the findings from this study as an important piece of the debate over whether AI will lead to cultural stagnation.

The results show that generative AI systems themselves tend toward homogenization when used autonomously and repeatedly. They even suggest that AI systems are currently operating in this way by default.

The familiar is the default

This experiment may appear beside the point: Most people don’t ask AI systems to endlessly describe and regenerate their own images. The convergence to a set of bland, stock images happened without retraining. No new data was added. Nothing was learned. The collapse emerged purely from repeated use.

But I think the setup of the experiment can be thought of as a diagnostic tool. It reveals what generative systems preserve when no one intervenes.

This has broader implications, because modern culture is increasingly influenced by exactly these kinds of pipelines. Images are summarized into text. Text is turned into images. Content is ranked, filtered and regenerated as it moves between words, images and videos. New articles on the web are now more likely to be written by AI than humans. Even when humans remain in the loop, they are often choosing from AI-generated options rather than starting from scratch.

The findings of this recent study show that the default behavior of these systems is to compress meaning toward what is most familiar, recognizable and easy to regenerate.

Cultural stagnation or acceleration?

For the past few years, skeptics have warned that generative AI could lead to cultural stagnation by flooding the web with synthetic content that future AI systems then train on. Over time, the argument goes, this recursive loop would narrow diversity and innovation.

Champions of the technology have pushed back, pointing out that fears of cultural decline accompany every new technology. Humans, they argue, will always be the final arbiter of creative decisions.

What has been missing from this debate is empirical evidence showing where homogenization actually begins.

The new study does not test retraining on AI-generated data. Instead, it shows something more fundamental: Homogenization happens before retraining even enters the picture. The content that generative AI systems naturally produce – when used autonomously and repeatedly – is already compressed and generic.

This reframes the stagnation argument. The risk is not only that future models might train on AI-generated content, but that AI-mediated culture is already being filtered in ways that favor the familiar, the describable and the conventional.

Retraining would amplify this effect. But it is not its source.

This is no moral panic

Skeptics are right about one thing: Culture has always adapted to new technologies. Photography did not kill painting. Film did not kill theater. Digital tools have enabled new forms of expression.

But those earlier technologies never forced culture to be endlessly reshaped across various mediums at a global scale. They did not summarize, regenerate and rank cultural products – news stories, songs, memes, academic papers, photographs or social media posts – millions of times per day, guided by the same built-in assumptions about what is “typical.”

The study shows that when meaning is forced through such pipelines repeatedly, diversity collapses not because of bad intentions, malicious design or corporate negligence, but because only certain kinds of meaning survive the text-to-image-to-text repeated conversions.

This does not mean cultural stagnation is inevitable. Human creativity is resilient. Institutions, subcultures and artists have always found ways to resist homogenization. But in my view, the findings of the study show that stagnation is a real risk – not a speculative fear – if generative systems are left to operate in their current iteration.

They also help clarify a common misconception about AI creativity: Producing endless variations is not the same as producing innovation. A system can generate millions of images while exploring only a tiny corner of cultural space.

In my own research on creative AI, I found that novelty requires designing AI systems with incentives to deviate from the norms. Without it, systems optimize for familiarity because familiarity is what they have learned best. The study reinforces this point empirically. Autonomy alone does not guarantee exploration. In some cases, it accelerates convergence.

This pattern already emerged in the real world: One study found that AI-generated lesson plans featured the same drift toward conventional, uninspiring content, underscoring that AI systems converge toward what’s typical rather than what’s unique or creative.

Lost in translation

Whenever you write a caption for an image, details will be lost. Likewise for generating an image from text. And this happens whether it’s being performed by a human or a machine.

In that sense, the convergence that took place is not a failure that’s unique to AI. It reflects a deeper property of bouncing from one medium to another. When meaning passes repeatedly through two different formats, only the most stable elements persist.

But by highlighting what survives during repeated translations between text and images, the authors are able to show that meaning is processed inside generative systems with a quiet pull toward the generic.

The implication is sobering: Even with human guidance – whether that means writing prompts, selecting outputs or refining results – these systems are still stripping away some details and amplifying others in ways that are oriented toward what’s “average.”

If generative AI is to enrich culture rather than flatten it, I think systems need to be designed in ways that resist convergence toward statistically average outputs. There can be rewards for deviation and support for less common and less mainstream forms of expression.

The study makes one thing clear: Absent these interventions, generative AI will continue to drift toward mediocre and uninspired content.

Cultural stagnation is no longer speculation. It’s already happening.

Ahmed Elgammal does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

– ref. AI-induced cultural stagnation is no longer speculation − it’s already happening – https://theconversation.com/ai-induced-cultural-stagnation-is-no-longer-speculation-its-already-happening-272488