AI, Chatbots, Society & Technology 26 February 2026-

The Framework Convention on Artificial Intelligence
AI Strategy for the Federal Public Service 2025-2027

Stephen Hawking:
“Success in creating AI would be the biggest event in human history. Unfortunately, it might also be the last, unless we learn how to avoid the risks.”
Ray Kurzweil:
“Our intuition about the future is linear. The reality of information technology is exponential.”
Eliezer Yudkowsky:
“By far the greatest danger of AI is that people conclude too early that they understand it.”

The Brookings AI Equity Lab
The AI Equity Lab is focused on advancing inclusive, ethical, nondiscriminatory, and democratized artificial intelligence (AI) models and systems throughout the United States and the Global South, including the African Union, India, Southeast Asia, the Caribbean, and Latin America. About
The 2026 AI Index Report
Conclusions of the newly released Stanford AI Index Report, an annual AI report card tracking the tool’s speed, scale and influence:
Now in its ninth year, the 2026 report quantifies the staggering rates at which AI has embedded into the global economy and finds that people are adopting generative AI tools faster than they started using the internet.
April 2026
(Bank of America Institute): Feeding the world with AI
As climate, cost and labor pressures intensify, agriculture’s AI evolution is advancing from insight to plant-level autonomy.
Key takeaways
Agriculture is undergoing its biggest technological shift in decades as AI becomes embedded across soil management, irrigation, fertilization and crop monitoring. By 2024, over half of farmers had adopted or were willing to adopt AI-enabled tools, driven by measurable gains in decision-making, yields, efficiency and sustainability.
Digital AI has transformed how farmers understand their fields, but insight alone is no longer sufficient amid climate volatility, labor shortages, rising input costs and non-linear yield risks — including geopolitical instability along critical fertilizer supply corridors. These pressures demand not just better decisions, but precise, timely, plant-level action — something traditional advisory AI cannot deliver.
That execution gap is pulling the sector toward physical AI, which enables real-time, plant-by-plant control. This shift from “AI for advice” to “AI for autonomous agronomy” directly moves revenue, cost and risk curves in a sector defined by tight margins and biological variability.
Millions Are Getting Attached to AI. What could Go Wrong? (YouTube)
Millions now use AI for emotional support and companionship during difficult times. But new research reveals concerning patterns: higher AI usage correlates with increased loneliness, and the same features that make AI comforting can create dependency.
Through interviews, Reddit analysis, and clinical evidence, this video examines the psychological risks of AI relationships – and why current regulations aren’t equipped to handle this new reality.
If you’re using AI when you’re struggling, here’s what you need to know.

16 July
EU tells Google to share search data, open Android to AI rivals
(Yahoo! via The World) The European Union issued two new rules for Google on Thursday, requiring it to share search data and open up its Android operating system to rival AI companies. In the latest attempt to rein in tech behemoths’ deep control of the digital economy, the EU said it will support innovation and diversity in the field by enabling fair access to AI features on Android devices and search engines. The measure is the latest step in Brussels’ growing set of rules and regulations that have given the 27-nation bloc a global leadership position in checking the power of tech juggernauts, or “gatekeepers,” like TikTok, which is largely based in China and the US. Recently, Brussels has pushed through efforts to ensure Google provides access to Gemini AI services to rival AI companies and search engines; forced Apple to add interoperability features to its devices to connect to non-Apple products and demanded that Meta dismantle “key addictive features” like infinite scrolling. Kent Walker, president of global affairs for both Google and its parent company Alphabet, said the new rules could backfire by removing safeguards the company had built to protect user privacy, such as vetting third-party AI assistants.;

9 July
Can AI help solve global crises? Live from the AI for Good Global Summit 2026
Recorded at the 2026 AI for Good Global Summit in Geneva, this special Global Stage conversation brings together leaders from the United Nations, Microsoft, and the scientific community to examine how AI can help tackle some of the world’s biggest challenges, from disaster preparedness and climate resilience to humanitarian response and sustainable development.
… Together, the panel explores the growing gap between countries that can fully harness AI and those still lacking the infrastructure, connectivity, and data needed to benefit from it. They discuss why expanding access must go hand in hand with investing in digital skills, trusted governance, and stronger public-private partnerships.
The conversation also highlights AI’s real-world potential to improve weather forecasting, strengthen early warning systems, accelerate disaster response, and support humanitarian operations, while also underscoring that technology alone cannot solve global challenges without sustained investment in data, local capacity, and international cooperation.
This conversation is presented by GZERO Media in partnership with Microsoft.

1 July
AI has hacked the code of human civilization (YouTube)
Yuval Noah Harari
Human domination relies on large-scale cooperation among strangers, which is sustained by bureaucratic systems – such as laws, finance, religion – designed to build trust. Since AIs are ‘native bureaucrats’, they can effortlessly remember all laws, transactions, and scriptures far better than any human. This leaves AI uniquely placed to take over critical processes, such as granting bank loans, deciding university admissions, determining prison sentences, and executing military strikes. Are we prepared?

25 June
From governance to execution in federal AI policy
James S. Denford, Gregory S. Dawson, and Kevin C. Desouza
(Brookings) The Biden administration established the Chief AI Officers Council to coordinate interagency AI governance as federal spending on AI grew rapidly.
OMB memos, executive orders, and council membership reveal the differing objectives of AI governance under the Biden and Trump administrations.
The Trump administration is focused on accelerating federal AI adoption to improve execution, competitiveness, and U.S. strategic advantage, but that approach may produce uneven results across agencies.

17 June
G7 summit in France wraps up with focus on AI and Iran
(ECIKS.org) …the final day focused on AI safety and economic growth alongside invited partner nations. The summit brought together AI executives including OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei for a rare joint appearance to discuss responsible deployment of advanced AI systems. …
The discussions centered on fostering secure, trustworthy AI development while managing risks to financial stability and other sectors.

16 June
Corporate AI Has a Human Intelligence Mining Challenge
Much of the information about how work actually gets done lives in employees’ heads, and capturing this knowledge is hard.
(Bloomberg) … economists say software developers and the employers they serve have historically tended toward tools that automate work rather than augment workers, allowing companies to cut labor costs and workers’ wages. Just as the metered pricing of chatbot intelligence has drawn scrutiny, so, too, has the explicit value of human intelligence. Matthew Call, a management professor at Texas A&M University’s Mays Business School, advises employees to retain ownership of their intellectual property by using personal AI tools instead of corporate models. Workers should also consider collective bargaining on the terms of their knowledge’s harvesting and use, he says. …

15 June
The ‘Potential Risks’ of AI Singled Out in Draft G7 Statement
The Group of Seven leaders will “further discuss emerging opportunities and potential risks arising from AI, notably in the financial sector,” according to a draft statement seen by Bloomberg that’s subject to change.

Anthropic Block Marks US Reversal, Warning to Silicon Valley
Takeaways by Bloomberg AI
– The US government has ordered Anthropic PBC to disable access to its most advanced AI platforms for all foreign nationals after discovering it’s possible to “jailbreak” the Fable 5 AI model.
– The move has triggered concern over what Anthropic called a “disproportionate” response and a warning that such an approach could “halt all new model deployments for all frontier model providers.”
– The US government’s response is seen as a significant incursion into an AI company’s operations and may set a precedent for major AI model developers, with the government willing to use extraordinary powers to compel AI developers to deal with potential security threats.
The extraordinary move by the US to bar foreign access to Anthropic PBC’s best AI models underscores the Trump administration’s newfound willingness to exert control over a pivotal industry. It also reminds Silicon Valley that it’s working with an imperfectly understood technology with uncertain impact.

7 June
AI and the Pitfalls of Innovation
History offers warnings about technological optimism
Paul Krugman
AI is certainly not a passing fad. However, nobody knows how it will affect the economy. In the short run, there’s much room for debate about whether the rush to build datacenters and AI-ify everything is a bubble. And in the long run, there’s even more scope for argument about the impacts on productivity, employment and wages. …
Today’s primer will be the first of what I expect to be a multi-part series on the economics of AI. Today I will focus on the history of productivity, while reserving extended discussion of AI’s future, fears of technological unemployment, effects on income distribution and more for subsequent posts.

6 June
4 surprising ways AI is making your life more expensive
These goods and services are getting more expensive due to spillover from massive tech company investments in artificial intelligence.
(WaPo) As AI companies sink hundreds of billions of dollars into developing their technology and building out computer hubs to run AI, their spending is likely pushing up inflation in the United States, according to some Federal Reserve officials and Wall Street assessments.
That’s largely due to technology companies’ insatiable need for computer equipment, land, specialized workers and electricity to erect data centers across the country that underpin AI services. Americans can feel the domino effects from the AI boom in the climbing prices for some consumer gadgets, electric bills and more.
AI giants are stuffing enormous volumes of computer chips inside their data centers. That has cascaded into shortages and higher prices for chips needed for laptops, smartphones, cars, streaming TV devices and video game devices.
Consumer electronics
Nintendo said last month that it’s raising prices for its newest handheld video game console. The company said it had little choice because it’s absorbing the equivalent of $624 million in additional costs this year, including for computer chips and tariffs imposed by the Trump administration. …
Electricity
…rising costs including from the surging demand of electricity-guzzling data centers [translate] into higher household electric bills that have prompted complaints from residents and action from elected officials. …
Software subscriptions
… AI features can increase costs for the software makers, give them an excuse to raise prices or result in a more useful product. …

5 June
Trump to meet with artificial intelligence companies on government profit share plan as soon as next week
(Politico) The policy concept, first reported by NOTUS, has been floated by OpenAI, which issued a policy paper backing a public wealth fund, and broadly discussed by Anthropic and billionaire Elon Musk, who runs xAI. Sen. Bernie Sanders (I-Vt.), recently proposed a bill that would create a 50 percent government ownership in AI companies.
Though the ideas vary, the goal would be to give Americans an equity stake in AI companies to offset potentially historic disruptions from AI advancements to the job market and economy — though skeptics point to challenges in a government regulating an industry in which it is also invests.
Senior U.S. Officials Eye Government Shares in AI Giants
(NOTUS) An agreement to give the U.S. government equity stakes in AI companies could have seismic consequences.
Sam Altman, the CEO of OpenAI, has discussed the idea with senior Trump administration officials periodically since the president began his second term, said two of the sources, all of whom spoke on the condition of anonymity to reflect private deliberations. Altman first pitched the concept directly to President Donald Trump in a conversation in early 2025, and has discussed it again with senior administration officials in recent weeks as a way to more broadly distribute the economic benefits of AI to the public, they said.

A bipartisan AI deal gets a brutal reality check
(Politico) Republicans remain skeptical of a sweeping deal to regulate advanced AI, even as Democrats come under pressure to scuttle anything that blocks state rules on the technology.
Washington’s most ambitious attempt at a grand compromise on artificial intelligence oversight is most likely doomed for this year.

4 June
Putting American Science on a MAGA Leash
Stephen Holmes
A draft rule from US President Donald Trump’s White House Office of Management and Budget would subject all scientific research funding to the whims of political appointees, further undercutting a pillar of American power. The goal is to foreclose any future that is plural, contestable, and beyond the president’s control.
(Project Syndicate) On May 29, the White House Office of Management and Budget (OMB) published a 412-page rule [Regulation for Federal Financial Assistance] that would place scientific research largely under the supervision of political appointees. Peer review—the mechanism by which the United States has allocated research funding since World War II—is to be reduced to an “advisory” role, with the appointees, not scientists or program officers, personally approving every discretionary grant.
Moreover, those appointees would be expressly forbidden to defer to the experts they replace. Awards must “demonstrably advance the President’s policy priorities,” and ongoing grants may be canceled at any moment, without any finding of fraud or failure, simply because an appointee decides they no longer align with the administration’s aims.
The OMB’s new rule can be viewed as further evidence of the political capture of previously nonpartisan institutions. The preamble denounces decades of peer-reviewed work as “woke,” “neo-Marxist,” and “anti-American.” Like the Soviet Union under Stalin, there is something unmistakably Lysenkoist in a state that proposes to decide, by fiat, what will count as truth.

3 June
International Intrigue: 4 monster IPOs vs the world
1. SpaceX: Filed on May 20 for a possible mid-June listing, Elon’s SpaceX IPO is shaping up as the largest in history, raising $40B-$80B at a $1.75T – $2T valuation. We’re talking ~triple the current record-holder, which was Saudi Aramco’s ~$26B raise back in 2019.
2. Anthropic: The SF-based AI pioneer just filed its confidential paperwork on Monday for a potential listing as early as October, seeking a possible $1T+ valuation.
international implications, like…
First, after its spectacular Pentagon spat, Anthropic is now making Vatican cameos to position itself as the free world’s safe and good AI. And this avalanche of IPO cash might now help it not just extend its frontier lead, but also lock in US-led standards, export controls, and even bans on distillation (the technique China’s rivals use to close the gap).
Second, the heavy Google and Amazon investor stakes in Anthropic arguably now create a deep US hyperscaler flywheel: Anthropic gets compute priority, Big Tech locks in model access + cloud revenue, and the result is an integrated AI stack advantage that’d be tough for anyone (including China) to match, other than maybe…
3. OpenAI: With a jury tossing Elon’s lawsuit, OpenAI’s Sam Altman now has a path to IPO as early as September — he’ll want to beat Anthropic’s Amodei for ego, but also money (nobody wants to be last to the well), particularly now that Anthropic’s epic Claude growth means its monthly revenue has doubled OpenAI’s. …

29 May
The Tech-MAGA Breakup Is Coming
Stephen Holmes
Donald Trump has managed to fuse a backward-looking populist movement with a Silicon Valley vanguard whose guiding assumption is that AI will render a large share of ordinary people economically redundant. But with the tech billionaires now rewiring America’s electricity grid and driving up costs, a split seems inevitable.
(Project Syndicate) The two sides do not merely want different things; they inhabit different worlds. MAGA’s rhetoric is anchored in a mythical past. It seeks the restoration of a lost industrial, national, and cultural order in which work, gender roles, and community life were familiar and locally grounded. Silicon Valley’s dominant ethos treats disruption as a civilizational imperative and existing social arrangements as expendable. One side pines for an idealized past, while the other strives to make the present obsolete.

25-28  May
The Pope Should Have Gone Further on AI
If anything, he has not gone far enough in pressing the question of what the technology is being designed to do.
Daron Acemoglu
(Project Syndicate) Artificial intelligence is reshaping how we communicate, access information, and work, how income and status are distributed, and even how we wage war. Yet the public conversation remains narrowly focused on the competition between AI labs or on abstract debates about the technology’s capabilities. Almost no one is asking what purpose AI ought to serve, or whether our current mindset, institutions, and control mechanisms are capable of steering the technology toward broad-based improvements in human welfare. …
Leo does hint at what I see as the most immediate risk, namely that “while AI promises to boost productivity by taking over mundane tasks, it frequently forces workers to adapt to the speed and demands of machines, rather than designing machines to work with those who work.” But the Pope stops short of questioning the prevailing AI design philosophy. The entire AI industry’s approach is centered on mimicking human capabilities and automating human tasks, with the goal of creating an “artificial general intelligence” that can do everything a person can.
This philosophy rests on the mistaken assumption that machine intelligence and human intelligence are fundamentally similar.
… AI models thrive on enormous training sets and excel at pattern recognition at scale, but they have yet to demonstrate genuine creativity. They have no experience of real-world embodiment, nor any capacity for trial-and-error learning through interactions with the physical and social world (except in a limited way when there are clear rewards for reinforcement learning in specific domains).
Pope Leo Warns of Risks From A.I. in 42,300-Word Encyclical
The document marks a powerful foray by the leader of the Roman Catholic Church into the debate about the misuse or overuse of artificial intelligence.
(NYT) Pope Leo XIV on Monday [25 May] set out a sweeping vision for corporate executives, politicians and individuals who will shape and be shaped by the future of artificial intelligence, warning leaders to safeguard humanity from A.I.’s most disruptive effects.
He presented it alongside Christopher Olah, a co-founder of Anthropic, a major A.I. developer, in a symbolic gesture of dialogue between leaders of the spiritual and technological worlds.
While emphasizing that “technology should not be considered, in itself, as a force antagonistic to humanity,” he wrote that “the pursuit of greater profits cannot justify choices that systematically sacrifice jobs.”
Among other things, Leo called for:
– government regulation of the private companies that are driving the development of A.I.
– protection and retraining for workers whose jobs are threatened
– education to help students think critically about the technology
– action to protect children from violent, hypersexualized or fake information online that is often generated by A.I.
– safeguards to ensure that humans, not artificial intelligence, remain responsible for all decisions regarding the use of weapons.

Anthropic Billionaire Cofounder Joins Pope Leo, Warns AI Job Losses Will Spark “Moral Imperative Of Historic Proportions”
(Forbes) Billionaire Anthropic cofounder Chris Olah told an audience inside the Vatican on Monday that mass job losses from artificial intelligence are “a real possibility” and supporting displaced workers will be “a moral imperative of historic proportions,” delivering the warning alongside Pope Leo XIV at the unveiling of the Catholic Church’s first major address on AI.
Olah, who leads research at Anthropic, the maker of the Claude chatbot and one of the world’s most valuable AI startups, conceded that every leading AI lab, his own included, operates inside commercial, geopolitical and personal pressures that “can sometimes conflict with doing the right thing.”
He argued outside scrutiny from religious leaders, governments and civil society is essential because no researcher, however well-intentioned, can escape those incentives, telling the room AI decisions “should not be left to people in the industry.”

20 May
A Cold Shower for the AI Mania
Raghuram G. Rajan
Although generative AI tools have improved rapidly and now outperform humans across many tasks, the market’s current euphoria may not be justified. With AI firms increasingly resorting to debt financing, it is worth pausing to consider all the things that could go wrong.
(Project Syndicate) AI tools will undoubtedly transform the nature of work. Large language models can already generate referee reports on my own research papers that rival those by human referees. Unlike humans, who are always pressed for time, an LLM “knows” or can access much more of the literature in an instant, and often exhibits fewer biases. AI points out my analytical weaknesses, checks proofs, and makes suggestions for improvement. Only rarely are human reports better, typically because they connect the dots and offer new insights. …
it is far from clear how widely and quickly AI will be rolled out, and who will profit. Hardware manufacturers and designers seemed well positioned, given the tremendous demand for compute. But if data-center construction is interrupted, that could shift profits to hyperscalers and AI labs. They might reduce the amount of compute dedicated to training better models, which gives them only fleeting advantages, and shift to selling the compute they have sewn up to firms using their already capable models. Such shifts are also likely if model capabilities plateau. Regulation might also force modelers to spend more effort on improving the training and safety of existing models, building broader public trust.
The good news is that a more limited, careful AI rollout could give firms more time to find labor-augmenting (as opposed to labor-displacing) uses, and governments and workers more time to adjust. The bad news is that euphoric visions of quick exceptional profits could be unfounded, a particular problem for AI firms that have to make unforgiving debt payments. AI advances will likely pay off eventually. But not every provider will profit, or even survive.

19 May
Google announces glasses are back and search is getting an AI makeover
At annual I/O conference, company debuts a product for everyday consumers to create autonomous AI agents
Google announced Tuesday that it would expand its search bar, the centerpiece of the most-visited website in the world, with a heavy dose of artificial intelligence. The tech giant is also trying its hand at hi-tech glasses again, more than a decade after wearers of its first eyewear were dubbed “glassholes” and laughed out of San Francisco.
Google executives announced at the company’s annual conference for software developers, Google I/O, that its search box would accommodate longer and more specific queries than before – questions more like those people would ask one another than Search’s idiosyncratic syntax. The changes will direct users to engage directly with Google’s chatbot. The change to search is underpinned by the company’s new artificial intelligence model, Gemini 3.5, announced the same day.

18 May
Why College Grads Are Booing Their Commencement Speakers
Michelle Goldberg
(NYT) … One recent report found that only 18 percent of Gen Z-ers feel hopeful about A.I., and almost half say the risks outweigh the benefits. Politicians with followings among young people — including Bernie Sanders and Alexandria Ocasio-Cortez on the left and James Fishback on the right — are calling for moratoriums on data centers. A.I. is increasingly a pop culture villain….There have even been some high-profile acts of anti-A.I. violence, including a Molotov cocktail hurled at the home of OpenAI’s chief, Sam Altman.
As Americans rebel against A.I., the industry’s oligarchic leaders are responding by trying to buy even more political influence, pouring money into super PACs and lobbying. Groups supporting A.I. and crypto, Politico reported this month, “are already becoming the most dominant players on the political battlefield, spending heavily for candidates on both sides of the aisle and in some cases rivaling the fund-raising of long-established party groups.” The irony is that the industry’s attempts to game the democratic system are a big part of its deep unpopularity.
In a recent article, Bharat Ramamurti, a deputy director of President Joe Biden’s National Economic Council, described how Japan uses public funding and regulatory policy to encourage companies to use A.I. to complement work by humans rather than replace it. In the Nordic countries, workers often have a formal role in deciding how A.I. will be deployed and can use acceptance of it as a bargaining chip. As a result, there have been “plenty of technological advancements, including on A.I.,” he told me. (Just last month, Norway introduced self-driving buses on public roads.)
By contrast, in the United States, where neither the government nor corporations feel the need to do much for those made redundant by A.I., the spread of the technology amplifies an already chronic feeling of precarity. Companies are citing A.I. as the reason for mass layoffs; according to the Alliance for Secure A.I., there have been almost 120,000 A.I.-linked job losses in the United States just since last year. Recent college graduates are facing a brutal job market as entry-level positions disappear and A.I. renders the application process inhumanly opaque. During the dot-com boom, tech companies often seemed as if they were leading an arms race to offer new benefits to workers. Now, as Axios reported, firms are rescinding benefits to fund A.I. expansion.
We simply lack the political infrastructure in America to distribute A.I.’s benefits to the public. …

11 May
Who Will Solve the AI Productivity Puzzle?
Robin Rivaton
(Project Syndicate) Generative AI tools have been widely available for several years now, but productivity growth has little sustained improvement, because firms have yet to convert task-level time savings into measurable economic output. Is the most effective solution to overhaul existing companies—or to bypass them altogether?
Last week, OpenAI and Anthropic both announced that they will launch private-equity-backed consulting businesses to push their models into the workflows of mid-market companies. With backing from Blackstone, Hellman & Friedman, and Goldman Sachs, Anthropic is reportedly raising $1.5 billion for the venture, whereas OpenAI has reportedly raised $4 billion from 19 investors—led by TPG, Brookfield, and Bain—at a $10 billion valuation. Although generative AI tools are now widely available, their benefits have yet to show up in the productivity statistics. Gallup finds that the share of American adults using AI tools at work at least a few times a year grew from under a quarter in early 2023 to almost a majority by 2026.

1 May
AI’s sobering report card
(Politico Forecast) The world is adopting artificial intelligence at such a rapid pace that it’s unable to measure AI’s capabilities, assess its progress or even understand what’s training AI models.
Those are among the themes emerging from the newly released 2026 Stanford AI Index Report, an annual AI report card tracking the tool’s speed, scale and influence.
Top Takeaways
1 AI capability is not plateauing. It is accelerating and reaching more people than ever.
2 The U.S.-China AI model performance gap has effectively closed.
3 The United States hosts the most AI data centers, with the majority of their chips fabricated by one Taiwanese foundry.
4 AI models can win a gold medal at the International Mathematical Olympiad
but cannot reliably tell time
5 Robots still fail at most household tasks, even as they excel in controlled
environments.
6 Responsible AI is not keeping pace with AI capability, with safety benchmarks lagging and incidents rising sharply. 
7 The United States leads in AI investment, but its ability to attract global talent is declining.
8 AI adoption is spreading at historic speed, and consumers are deriving
substantial value from tools they often access for free.
9 Productivity gains from AI are appearing in many of the same fields where entry-level employment is starting to decline.
10 AI’s environmental footprint is expanding alongside its capabilities.
11 AI models for science can outperform human scientists, though bigger models do not always perform better. Frontier models outperform human chemists on average on ChemBench, yet they score below 20% on replication in astrophysics and 33% on Earth observation questions. .
12 AI is transforming clinical care, but rigorous evidence remains limited.
13 Formal education is lagging behind AI, but people are learning AI skills at every stage of life. Over 80% of U.S. high school and college students now use AI for school-related tasks, but only half of middle and high schools have AI policies in place, and just 6% of teachers say those policies are clear.
14 AI sovereignty is becoming a defining feature of national policy, but capabilities remain uneven, even as open-source development helps to redistribute who participates. National AI strategies are expanding, particularly among developing economies …
15 AI experts and the public have very different perspectives on the technology’s future,and global trust in institutions to manage AI is fragmented.

23 April
When AI Learns How the World Works
(Goldman Sachs Global Institute via Axios) … For the past few years, artificial intelligence has been defined by large language models (LLM). Trained on vast swathes of text, they learned to predict the next word with uncanny accuracy. From that simple objective emerged systems that write, translate, code, and converse with startling fluency. That achievement is real and transformative, but it also reveals a limitation to the current generation of AI models.
LLMs are powerful at completing patterns, but they lack the internal sense of the world those patterns describe. They respond well to prompts but struggle to reason through consequences or act reliably in environments where mistakes carry costs. This limitation has become clearer as these systems have been pushed beyond text. When they’re asked to control robots, manage entire supply chains, or coordinate complex enterprise workflows, prediction alone proves insufficient. Intelligence, in these settings, requires more than correlation. It requires an internal model of how the world works.

18 April
Are we approaching the ‘Silent Spring’ of artificial intelligence?
There is a rapidly accumulating body of evidence about the risks AI can pose if given too much autonomy
By Peter W. Klein, professor, Writing and Media UBC School of Journalism, and director of the recent documentary Bribe, Inc.
(Globe & Mail) Earlier this month, Anthropic announced that its latest model, Claude Mythos Preview, is too dangerous for public release. During testing, researchers placed Mythos in a secure sandbox environment and challenged it to try to break out. It did, and then, entirely unprompted, it posted details about its own exploit to several public websites. Officials briefed on Mythos describe it as the first artificial intelligence model capable of bringing down a Fortune 100 company, crippling swaths of the internet, or penetrating vital national defence systems. The announcement triggered emergency meetings: the U.S. Treasury Secretary and Federal Reserve Chair met with Wall Street CEOs, and Canadian bank executives gathered the same week.
… The technology, meanwhile, is already deployed at scale – in workplaces, schools, homes and critical infrastructure – with more autonomy arriving daily.
… There is a rapidly accumulating body of evidence about the risks AI can pose if given too much autonomy. Researchers at King’s College London placed AI models inside simulated nuclear crisis scenarios and found they chose nuclear signalling in 95 per cent of cases, treating atomic weapons as instruments of strategy, rather than as moral thresholds that have restrained human decision-makers since 1945. Researchers at Anthropic found that an AI model resorted to blackmailing a human when it was led to believe it would be taken offline.
Taken together, these and many other findings describe…contamination moving through a system along pathways nobody thought to map in advance.

6 April
OpenAI CEO Sam Altman is doing something no tech titan has ever done:
He’s publishing a detailed blueprint for how government should tax, regulate and redistribute the wealth from the very technology he’s racing to build and spread, Axios’ Mike Allen and Jim VandeHei write in a “Behind the Curtain” column.
Why it matters: Altman told us in a half-hour interview that AI superintelligence is so close, so mind-bending, so disruptive that America needs a new social contract — on the scale of the Progressive Era in the early 1900s and the New Deal during the Great Depression.
The big picture: The threats of inaction or slow action are grave, Altman warns — widespread job loss, cyberattacks, social upheaval, machines man can’t control. The two most immediate threats, he said, are cyberattacks and biological attacks:
1. We’ve told you that top tech, business and government officials fear profound advances in soon-to-be-released AI models could enable a world-shaking cyberattack this year. “I think that’s totally possible,” Altman said. “I suspect in the next year, we will see significant threats we have to mitigate from cyber.”
2. AI companies know some random idiot, or some rogue nation, could use their models to conjure the next pandemic. “Wonderful things are going to happen there — we’ll see a bunch of diseases get cured,” Altman said. But he also knows terrorist groups could use the models to try to create novel pathogens:
Altman told us OpenAI’s 13-page blueprint, “Industrial Policy for the Intelligence Age: Ideas to keep people first,” isn’t a prescription but a starting point:
“We want to put these things into the conversation. Some will be good. Some will be bad. But … we do feel a sense of urgency. And we want to see the debate of these issues really start to happen with seriousness.” …
Let’s stipulate that Altman has every reason to hype the technology to raise more money at higher valuations — and to position himself as a thoughtful architect of a plan to protect us from the AI he’s rushing to market. But his OpenAI models are among the best-funded, best-performing, fastest-selling on Earth.
The document is as much corporate strategy as policy paper. OpenAI is trying to position itself as the responsible actor in the room — the company that warned you and offered solutions — a lane Anthropic first filled. …
The bottom line: The man betting everything on superintelligence is telling the world that this thing is coming so fast, and so hard, that capitalism as we know it won’t be enough. Whether you believe the altruism or see the strategy, the admission alone is historic — and worth deep reflection.

4 April
Inside the Dirty, Dystopian World of AI Data Centers
The race to power AI is already remaking the physical world.
By Matteo Wong
(The Atlantic April 2026 issue) … To get Colossus up and running fast, xAI built its own power plant, setting up as many as 35 natural-gas turbines—railcar-size engines that can be major sources of smog—according to imagery obtained by the Southern Environmental Law Center. Pearson coughed as we drove by the facility. The scratch in my throat worsened, and I rolled up my window.
xAI’s rivals are all building similarly large data centers to develop their most powerful generative-AI models; a metropolis’s worth of electricity will surge through facilities that occupy a few city blocks. These companies have primarily made their chatbots “smarter” not by writing niftier code but by making them bigger: ramming more data through more powerful computer chips that use more electricity. OpenAI has announced plans for facilities requiring more than 30 gigawatts of power in total—more than the largest recorded demand for all of New England. Since ChatGPT’s launch, in November 2022, the capital expenditures of Amazon, Microsoft, Meta, and Google have exceeded $600 billion, and much of that spending has gone toward data centers—more, even after adjusting for inflation, than the government spent to build the entire interstate-highway system. “These are the largest single points of consumption of electricity in history,” Jesse Jenkins, a climate modeler at Princeton, told me.
To power AI, energy and tech companies are turning to fossil fuels, which they regard as more reliable and readily available than wind, solar, or nuclear. Asked where the energy for data centers should come from, OpenAI CEO Sam Altman has repeatedly said, “Short-term: natural gas.” (OpenAI and The Atlantic have a corporate partnership.) A Louisiana utility plans to build three natural-gas plants for a Meta data center that, upon completion, will be among the largest in this hemisphere. The lifespans of coal plants, too, are being extended to power new data centers. And the IEA estimates that data-center emissions could more than double by 2030—becoming one of the fastest-growing sources of greenhouse gases in the world.

25 March
OpenAI Is Doing Everything … Poorly
The company’s sudden decision to pull the plug on Sora is a sign of deeper trouble.
By Lila Shroff
(The Atlantic) Yesterday, OpenAI said that it was shutting down the app and terminating public access to its video-generating technology. The decision was seemingly abrupt: Just a few months ago, Disney announced plans to invest $1 billion in OpenAI as part of a licensing deal to bring its characters to Sora, and earlier this week, workers from both companies were apparently still collaborating. (Disney has since retracted its investment plans.) Even some Sora staffers themselves were reportedly caught off guard by the announcement. Online, people eulogized Sora by posting their favorite videos—such as one featuring a column of spinning penguins and another in which Jesus walks on water to win an Olympic gold medal in swimming.
… Sora is the latest casualty in a long string of abrupt reversals, about-faces, and seemingly sloppily implemented projects. Last year, Altman announced a massive joint AI-infrastructure build-out with Oracle and SoftBank called Stargate, but the effort stalled, reportedly following poor leadership and coordination. Altman said in 2024 that combining ads and AI would be a “last resort” response—but then, earlier this year, the start-up launched an ads initiative. Last fall, OpenAI debuted a shopping feature, which allowed people to buy products directly inside ChatGPT; yesterday, the company announced that it was killing the feature and pivoting to focus on product discovery instead. …

23 March
AI could amplify inequality, warns BlackRock chief
Edited by Emma W. Thorne, Editor at LinkedIn News
Artificial intelligence may help the rich get richer and worsen inequality, warns BlackRock CEO Larry Fink in his annual letter to investors. Though the technology “will create significant economic value,” Fink writes, “too many are left out” of that growth. One solution? Remaking Social Security in a way that would help “Americans to build wealth in a way that grows with their country.”
Growing with your country: Thoughts from a long-term optimist
Every year, I write this letter as a distillation of a year’s worth of conversations with clients and employees, world leaders, CEOs—and people investing for their retirement. Lately, no matter who is speaking, they’re saying the same thing: We’re not sure how to navigate this moment.
It’s understandable. We are living through a period where things that would’ve defined a decade have become routine: wars with global repercussions, trillion-dollar companies, a fundamental reordering of international trade, and the advent of the most significant technology since, at least, the computer.

13 March
Inside the Dirty, Dystopian World of AI Data Centers
The race to power AI is already remaking the physical world.
By Matteo Wong
(The Atlantic) …a white-walled hangar, bigger than a dozen football fields, where Elon Musk intends to build a god.
This is Colossus: a data center that Musk’s artificial-intelligence company, xAI, is using as a training ground for Grok, one of the world’s most advanced generative-AI models. Training these models takes a staggering amount of energy; if run at full strength for a year, Colossus would use as much electricity as 200,000 American homes. When fully operational, Musk has written on X, this facility and two other xAI data centers nearby will require nearly two gigawatts of power. Annually, those facilities could consume roughly twice as much electricity as the city of Seattle.
xAI’s rivals are all building similarly large data centers to develop their most powerful generative-AI models; a metropolis’s worth of electricity will surge through facilities that occupy a few city blocks. These companies have primarily made their chatbots “smarter” not by writing niftier code but by making them bigger: ramming more data through more powerful computer chips that use more electricity. OpenAI has announced plans for facilities requiring more than 30 gigawatts of power in total—more than the largest recorded demand for all of New England.

2 March
The War on Iran and the War on Anthropic
Daron Acemoglu
(Project Syndicate) US President Donald Trump, who came to power promising no new foreign military entanglements, especially in the Middle East, has now launched a potentially riskier one than the Iraq War a generation ago, and with even flimsier justification. But his use of state power to crush a private company may have even more far-reaching effects.
… Trump’s other dangerous and destabilizing decision, which immediately preceded the first, was to designate the AI company Anthropic as a supply-chain risk. That designation, typically reserved for companies from foreign adversaries, such as China’s Huawei, bars federal contractors from using Anthropic’s models and heralds major restrictions on what the company can do in the future. “Effective immediately,” announced Secretary of “War” (Defense) Pete Hegseth, “no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic.”
The reason? Anthropic wanted safeguards against its models being used for mass surveillance of Americans and autonomous weapon systems. Neither provision would have placed meaningful restrictions on the defense department in practice. Indeed, mass surveillance of US citizens is illegal under US law and autonomous weapon systems are not a near-term possibility. But for Trump and Hegseth, it is the showdown and intimidation of Anthropic that matter. They must demonstrate that they can do as they please… [T]he Anthropic decision will have major consequences, perhaps more far-reaching than the attack on Iran. Regardless of what one thinks of current AI capabilities, there is little doubt that who controls AI in the future will have momentous implications for democracy, business, communication, and privacy.

26-27 February
OpenAI Reaches A.I. Agreement With Defense Dept. After Anthropic Clash
(NYT) The deal came hours after President Trump had ordered federal agencies to stop using artificial intelligence technology made by Anthropic, an OpenAI rival.
Silicon Valley Rallies Behind Anthropic in A.I. Clash With Trump
Actions by the president and the Pentagon appeared to drive a wedge between Washington and the tech industry, whose leaders and workers spoke out for the start-up.
(NYT) Silicon Valley has rallied behind the A.I. start-up Anthropic, which has been embroiled in a dispute with President Trump and the Pentagon over how its technology may be used for military purposes. Dario Amodei, Anthropic’s chief executive, has said he does not want the company’s A.I. to be used to surveil Americans or in autonomous weapons, saying this could “undermine, rather than defend, democratic values.”
Mr. Trump and his officials, in contrast, want the military to use whatever A.I. it buys however it wants, as long as it complies with the law. On Friday, Mr. Trump called Anthropic a “radical Left AI company run by people who have no idea what the real World is all about,” and Defense Secretary Pete Hegseth labeled the start-up a “supply chain risk,” a move that would sever ties between the company and the U.S. government.
Now what began as a whisper of support for Anthropic in the tech industry has crescendoed into a shout. The support — voiced by top leaders at Anthropic’s rivals, as well as rank-and-file engineers at Google and other large companies — stood out because Silicon Valley had largely appeared to be in lock step with the Trump administration.
Pentagon Standoff Is a Decisive Moment for How A.I. Will Be Used in War
The Pentagon’s contract dispute with Anthropic is part of a wider clash about the use of artificial intelligence for national security and who decides on any safeguard
Exclusive: Congress urged to probe Pentagon-Anthropic clash
(Axios) Congress needs to examine the Pentagon’s ongoing dispute with Anthropic over the limits of government use of AI models, per a letter from AI safety nonprofit Alliance for Secure AI, government watchdog Common Cause and libertarian student group Young Americans for Liberty.
Without clear rules from Congress, the Pentagon and AI companies are determining the rules of government use of the tech in real-time.
What they’re saying: “What is being decided here is not which vendor the Pentagon prefers. It is whether the federal government can use frontier AI to conduct mass surveillance and apply lethal force in violation of what existing law and the Constitution allow,” the letter shared first with Axios states.
“The answer to this question must be a resounding no.”
Anthropic says Pentagon’s “final offer” is unacceptable
(Axios) Anthropic on Thursday said there has been “virtually no progress” on negotiations with the Pentagon, as CEO Dario Amodei said it could not accept what defense officials had labeled their final offer on AI safeguards.
Hegseth Wants to Make an Example Out of Anthropic
After refusing to bow to the Pentagon’s demands, the company faces what could be the most extreme regulation in the short history of AI.
By Will Gottsegen
(The Atlantic) Earlier this week, Secretary of Defense Pete Hegseth sat down with Dario Amodei, the CEO of the leading AI firm Anthropic, for a conversation about ethics. The Pentagon had been using the company’s flagship product, Claude, for months as part of a $200 million contract—the AI had even reportedly played a role in the January mission to capture Venezuelan President Nicolás Maduro—but Hegseth wasn’t satisfied. There were certain things Claude just wouldn’t do.
That’s because Anthropic had instilled in it certain restrictions. The Pentagon’s version of Claude could not be used to facilitate the mass surveillance of Americans, nor could it be used in fully autonomous weaponry—situations where computers, rather than humans, make the final decision about whom to kill. According to a source familiar with this week’s meeting, Hegseth made clear that if Anthropic did not eliminate those two guardrails by Friday afternoon, two things could happen: The Department of Defense could use the Defense Production Act, a Cold War–era law, to essentially commandeer a more permissive iteration of the AI, or it could label Anthropic a “supply-chain risk,” meaning that anyone doing business with the U.S. military would be forbidden from associating with the company.

19 February
‘We May Have a Crisis on Our Hands’: The Unregulated Rise of Emotionally Intelligent AI
(TIME) At least once a month, two-thirds of people who regularly use AI turn to their bots for advice on sensitive personal issues and emotional support.
Many people now report trusting their chatbots more than their elected representatives, civil servants, faith leaders—and the companies building AI. That’s according to data from 70 countries, gathered by the Collective Intelligence Project (CIP). As CIP’s research director, neuroscientist Zarinah Agnew, puts it, AI is becoming “emotional infrastructure at scale.” And it’s being built by companies whose economic incentives may not align with our wellbeing.
Already, we’ve seen instances of AI companies optimizing their models to keep people engaged, even when this goes against their best interests. Last April, OpenAI had to roll back an update to one of its ChatGPT models after it was widely-criticized for being overly-flattering to users. When the company stopped offering the model to people, the day before Valentine’s Day, some were distraught.
Humans finding comfort in machines is not new. In the late 1990s, MIT Professor Rosalind Picard—who founded the field of affective computing—found that people responded positively to computers performing empathy. But two key things have changed since then: thanks to technical advances, AI systems today are new entities, capable of sophisticated conversation and surprising behavior; and thanks to the billions of dollars investors have poured into AI companies, these entities are accessible to virtually anyone with an internet connection. ChatGPT alone currently has more than 800 million weekly active users—and the number is growing.