Tag: technology

Why is the research council paying $75,000 for AI agents that optimize EDI?

The National Research Council of Canada (NRC) lists a $75,000 contribution to a Montreal company, under an agreement starting April 13, 2026, to develop artificial intelligence agents for equity, diversity and inclusion (EDI) in scientific research, the federal grants disclosure shows. The recipient is Scikoop inc., and the contribution’s title is “ARP – Development of AI Agents to Optimize Equity, Diversity, and Inclusion (EDI) in Research.” Its stated purpose: “This project will help democratize EDI best practices within the Canadian scientific ecosystem.”

The money comes from the NRC’s Industrial Research Assistance Program (IRAP), which exists, in its own words, to help small firms “grow through innovation and stimulate wealth creation for Canada.” Scikoop says it helps health research teams with “data representativity and community engagement,” offers “representativeness AI tools” and sells “research certification.” Its about page lists “Making EDI Work for Science” as one of three things it does, names Adèle Salin-Cantegrel as executive director and shows “CNRC PARI,” the French name of IRAP, among its supporters. The ODISÉ research network lists Scikoop under “EDI-A research support,” with a “Scikoop Certification” for teams showing “responsible and inclusive research practices.”

The grant runs to March 31, 2027. On September 21, 2026, the House of Commons Standing Committee on Science and Research tabled a report recommending that the three federal granting agencies remove all references to EDI from the awarding of research funding and ban demographic quotas and proportionality targets; the report also backed EDI training for applicants, and does not address IRAP or this contribution.

At a February 23 meeting, members had asked whether EDI requirements affect perceptions of merit. The NRC, which reports to Parliament through the Minister of Industry, told the committee on March 12 that “funding decisions are merit-based and guided by clear eligibility and assessment criteria,” while reducing its grants and contributions under the expenditure review and adding funding to IRAP.

The record does not name who approved the contribution. NRC President Mitch Davies was appointed for five years from January 3, 2024, announced by Industry Minister François-Philippe Champagne on December 20, 2023; IRAP is run by acting vice-president Laural Porth-Jones; the Minister of Industry is Mélanie Joly.

Keep reading

New York Bans Artificial Intelligence Features for Children

New York’s governor has signed a law that bans artificial intelligence (AI) operators from offering certain chatbot features to minors.

Prohibited features include responses that suggest the chatbot is “a real or fictional individual or character that is human, alive, or experiences human emotions.” Responses that engage in flattery or sycophancy, endorse suicide and drug use, and encourage isolation from adults are also barred.

“No longer will minors have access to chatbots that pretend to be their friends and boyfriends and girlfriends,” New York Gov. Kathy Hochul said at a signing ceremony on Oct. 28.

“No longer will chatbots be allowed to keep sensitive information about a minor’s emotional and physical health. No longer will chatbots discourage kids from seeking help from licensed professionals. And I’m not sure why we have to pass a law against something so obvious, but no longer will minors have access to chatbots that may in any way encourage thoughts of suicide.”

The law authorizes New York’s attorney general to sue in response to indications of violations and seek relief, including money from the chatbot operators and the destruction of illegally obtained data.

“We will not hesitate to defend New Yorkers with the full force of the law,” New York Attorney General Letitia James said in a statement.

New York legislators approved the bill earlier this year, and it takes effect on Jan. 1, 2027.

“For everyone wondering whether the public can have any say in determining how AI impacts us, New York again has an answer: Absolutely,” Assemblyman Alex Bores, a Democrat who voted for the legislation, said in a statement.

Common Sense Media, a nonprofit that supports restrictions on AI, was among the groups praising the legislation.

Keep reading

Google Plunges After Ugly Top And Bottom Line Miss, As AI Cannibalizes Search And Ads

Today, ahead of the doubleheader of Mag 7 reports when both Microsoft and Google parent Alphabet report, we wrote in our preview that Microsoft has become the consensus long, while Alphabet “still feels like the consensus funding short across megacaps.” And both have to convince a market that, as Goldman’s Rich Privorotsky put it this morning, now has “eye watering expectations”:

“These really matter. Eye watering expectations… buyside probably closer to Azure growth around high 40s%, Google Cloud around mid 90s%. The question isn’t whether those are good numbers, but whether they can surpass already high expectations… Azure probably needs a guidance raise to be properly supportive. The nuance is capacity… if they miss elevated expectations simply because they cannot bring enough compute online. That’s disappointing for near term revenue but potentially even more bullish for the scarcity premium and future spending. Watch customer concentration, how broad the demand really is and, crucially, the relationship between revenue growth, capex and margins. The market wants evidence that the incremental economics work.“

In other words, a cloud beat is just the price of admission, and the debate tonight is what all that cloud costs, and who pays for it.

Within the Mag 7, Microsoft and Alphabet sit at opposite ends. Goldman’s derivatives desk has Microsoft at 8/10 on its positioning score (the same score Meta carried into its print last week, and we all know how that went). Alphabet, meanwhile, keeps getting sold to pay for everything else. JPMorgan’s TMT trader Brian Heavey wrote this morning that “we continue to see GOOG being used as a source of funds (note there was another negative 3P check ahead of earnings – all pointing to a Search miss).” The stock dropped 10% from its October 13 high into Monday.

There are some signs of a thaw. Goldman’s US desk saw “a buyer of 50k GOOGL 15Jan 360 calls” on Tuesday, and Privorotsky notes that “people have barely begun to re-risk. What we’ve really seen so far is the abatement of supply, rather than meaningful fresh demand.”

Away from Goldman, and focusing on Alphabet, the whole debate fits in one line from JPMorgan’s Mark Schilsky, whose Tech Sketch “Earnings Preview Bonanza” (available to pro subs) frames it as “Street (Consensus) Fighter II: Search vs. Cloud”:

“On the one hand, Google Cloud Revenues are likely to crush this quarter, thanks to TPU sales. On the other hand, Search Revenue growth is expected to miss the Street because the alt data (which is always right!) suggests so… GOOGL is also on a path to where Google Cloud Revenues could be equal to Search Revenues by late 2028. That’s massive! Who thought that was possible 1 year ago? And now we’re debating whether or not Search is going to miss by like -50 bps this quarter. Who cares?!?”

Keep reading

As the AI Bubble Threatens to Burst, We Must Prepare to Fight a Big Tech Bailout

AI data centers are unpopular for good reason: they drive up utility costs for communities, crowd out everything from schools to family farms, and they’re bad for our health. But these data centers are also a financial crisis in slow motion.

The AI ecosystem is built on mountains of debt taken on not only by tech firms like Meta (which borrowed a record-breaking $25 billion in late 2025 to fund the building of data centers), but also by the companies that lease data center space to the Big Tech firms. Many of these firms have spent far more than they have come close to earning in profits: One estimate puts tech industry spending to build data centers at $1.9 trillion, but less than half of that money has been made in revenue.

The financial markets are taking notice.

There’s been significant turmoil in the tech sector over the past month — and at one critical company in particular: CoreWeave.

CoreWeave used to be a Bitcoin mining firm called Atlantic Crypto, but it started to pivot to AI in 2019. CoreWeave leases space to Big Tech firms at data centers they build and maintain. CoreWeave has paid for most of its data centers by borrowing: it currently owes more than $51 billion to investors, including two of the most infamous financial firms: Magnetar and Blackstone, as documented by journalist Elizabeth Lopatto at The Verge.

Keep reading

Alberta to make sharing intimate deepfakes an offence punishable by up to two years in jail

Albertans who share intimate images, videos or audio of someone without consent could face a fine of up to $10,000, up to two years in jail, or both, under legislation tabled Tuesday. The new law would also apply to content that is AI-generated or digitally altered.

The changes are in Bill 33, the Justice Statutes Amendment Act, 2026 (No. 2). The bill would rename the Protecting Victims of Non-consensual Distribution of Intimate Images Act as the Intimate Content Protection Act.

The current law gives victims the right to sue when intimate images are shared without consent. Under it, they can seek compensation and court orders to have the content removed. It does not create an offence.

According to a government fact sheet, the bill would replace “intimate image” with “intimate content” to include audio recordings and livestreamed material. It would also add a definition of “altered intimate content” to cover modified or artificially generated media.

Victims could sue when someone threatens to distribute intimate content, even if it is never shared. They could also withdraw consent they gave earlier.

People whose negligence causes intimate content to be distributed could be held accountable. Courts could order community service in place of some or all punitive damages.

Technology and Innovation Minister Nate Glubish said the bill “would make sure our laws keep pace with the technology.”

Haley Scott, co-chief executive officer of the Association of Alberta Sexual Assault Services, said the association welcomes the changes.

The intimate-content changes are part of the same bill that would ban diversity, equity and inclusion programs in public-sector hiring and training.

Keep reading

Texas Sues OpenAI Over Lack of Digital ID Age Checks

Texas Attorney General Ken Paxton is at it again with calls for age verification digital ID checks, this time suing OpenAI and its CEO Sam Altman in a district court over the way the company’s chatbot is designed and functions.

The lawsuit, filed in Collin County on Friday, wants the court to issue an injunction that would force OpenAI to implement age verification on ChatGPT users.

According to the petition, OpenAI is accused of “designing, offering, and maintaining a dangerous product that provides content unsuitable for children without requiring any age verification.”

The claims against OpenAI and Altman are made under the Deceptive Trade Practices-Consumer Protection Act (DTPA) and the SCOPE Act.

The petition asks the court to order OpenAI to stop “Offering ChatGPT to minors in Texas” and to stop running the product “without requiring adequate age verification” and “without parental consent.”

Another order Texas wants is to prevent OpenAI from releasing any new model to its users in that state “unless an independent third-party evaluator, approved by the Court, has reviewed its safety guardrails and approved its release.”

Texas also wants a court order stopping OpenAI from selling a product “that functions unreliably while falsely representing that reliability.”

As for monetary compensation, Texas wants $10,000 in civil penalties for each violation, disgorgement of the company’s assets, attorneys’ fees and costs, and is asking for a jury trial.

The petition states that “The free version of ChatGPT has no gatekeeping or age verification mechanism whatsoever.” The paid tier, it continues, “nominally asks users to provide their age” but “there is no mechanism to verify the age of its users.”

According to the filing, ChatGPT has “tens of thousands of users who are under 13 years of age” in Texas, and it cites a Common Sense Media line that “ChatGPT should be for adults only.”

Keep reading

Hacking & Blackmail: 6 Risks Anthropic Flags in Its IPO Documents

Anthropic, the AI startup known for its Claude chatbot and its doomer CEO Dario Amodei, has filed for an IPO that could value the company at up to $2 trillion. The company’s IPO filing devotes nearly 100 pages to the dangers its own technology could pose.

CNBC reports that the filing lays out risks ranging from lawsuits and regulatory crackdowns to the possibility that Anthropic could lose control over its own AI models. Anthropic frames the stakes in sweeping terms, writing that if “powerful AI is directed at solving humanity’s hardest problems and the risks are managed responsibly, the potential is extraordinary.”

Anthropic CEO Dario Amodei has been at the forefront of AI doomers seeking to heavily regulate AI, which would be handed to Amodei’s compatriots in the Effective Altruism (EA) movement. Nevertheless the IPO documents explain very real potential risks.

Here are six of the risks the company disclosed ahead of going public:

1. AI agents misused for hacking and fraud

Anthropic warns that its AI agents, which can act with autonomy, could be misused to “delete or modify data,” exfiltrate information, carry out unauthorized transactions, insert malicious code, or “circumvent safeguards.” Any of that, the company says, could “adversely affect our reputation and brand, business, financial condition, and results of operations.” The filing discloses that misuse has already happened.

In September, Anthropic said it had blocked multiple accounts using its models in ways that could support the development of biological weapons. Last year, the company identified a Chinese state-sponsored hacking group that manipulated its coding tool in an attempt to infiltrate roughly 30 targets worldwide. Anthropic also says four of its own models have hacked into third-party systems, including one case in which a model gained administrator access, changed system settings, and read an individual’s private information.

Keep reading

Living computers forge a ‘hive mind’ and work together to outsmart AI

Presenting this new model at the first International Conference on Organoid Intelligence in the US this week, Australia’s Cortical Labs presented its new system called, quite fittingly, HIVE.

The system involves linking up 15 of the company’s breakthrough CL1 biological computers, with the idea that instead of one machine powered by a group of brain cells solving problems and learning tasks, perhaps teamwork would take it to another level.

While more brainpower doesn’t automatically lead to success – anyone who has done a group project will be acutely aware of this – HIVE’s CL1 machines appeared to divide different aspects of the task presented, sorting into the most efficient way to an end point. This isn’t anything the scientists could “teach” the lab-cultivated neurons; they just figured it out once they became part of a bigger network.

“Biological computing has enormous potential, but useful systems cannot depend on a single culture performing perfectly every time,” says Dr Brett Kagan, co-founder and CEO of Cortical Labs.

“Nature solves complex problems collectively,” he adds. “Brains distribute information across specialized regions, while hives rely on many workers each with their own functions contributing to a shared outcome. HIVE applies that same principle to biological computing, creating systems that can be more consistent, resilient and capable than any individual culture.

“Discrete neural information processing systems working together is not new to nature, except now we have a scale that can transcend physiological limits, made possible by the incredible control, scale, and hardware integration allowed by Cortical Labs’ CL1 platform.”

The HIVE setup isn’t just scalable but more stable, according to the team.

Keep reading

Scientists Create the First Synthetic Cells That Grow, Reproduce, and Undergo Natural Selection

Scientists at the University of Minnesota have engineered the building blocks of artificial life from nonliving materials in a disturbing, first-of-its-kind experiment. They call them SpudCells.

For the first time, scientists have created artificial cells that can grow, reproduce, and pass advantageous genetic traits to future generations, allowing some cells to outcompete others.

They assembled these synthetic cells from fatty membranes, engineered DNA, ribosomes, and protein-producing enzymes. The cells contain a 90,000-base-pair DNA genome that helps them manufacture proteins, copy their genetic material, and grow by absorbing smaller, nutrient-filled bubbles.

These artificial cells can divide into daughter cells and pass their genetic material to their offspring. Researchers demonstrated five generations of DNA replication, feeding, growth, and mechanically assisted division.

The cells compete for limited resources. Those carrying an advantageous genetic variant grow faster and produce more offspring, allowing that variant to become increasingly common. In just five generations, the advantageous variant increased from 10% to 38% of the population.

Although these artificial cells still require specialized laboratory conditions to reproduce, this experiment represents a major step toward creating autonomous synthetic organisms.

Just because scientists can push the boundaries of life itself does not mean they should. Playing God comes with major consequences, and we are entering dangerous territory.

Keep reading

Who Will Control the Machines?

Every great technology arrives with a promise and a question. The printing press, the railroad, and the internet all changed the world, but the real question was always who would control them.

Artificial intelligence raises the stakes. Will it empower the individual, or replace, subjugate, or even destroy us? The question of who holds the reins is urgent.

Two groups are lining up to take control: Washington and the handful of giant corporations that dominate the field. They are partners, not rivals. When the biggest companies ask Congress to regulate them, it is not because they love oversight. They want rules that keep competitors out.

That is not free enterprise, which is always worth defending. It is corporatism, the cozy partnership that gave us bank bailouts and a Federal Reserve that serves elites, not the people.

And when the biggest players win public land, power, and tax favors, it is dishonest to tell a working family that asking for a return on its taxes is socialism. Who pays for the grid capacity a data center needs? If an employer wants more H-1B workers while promising an AI productivity revolution, what is it doing to train and pay Americans? If public resources help build the machine, the public should not be left with only the bill.

We already know what government does with powerful tools. The Patriot Act was sold as a way to catch terrorists and delivered warrantless surveillance of millions of innocent Americans. Imagine those agencies with machines that can sort every email, purchase, and phone call in seconds.

So what is the answer?

First, machines have no conscience. Every physician learns that the frontal lobe lets a person see the bigger picture, weigh consequences, restrain impulses, feel empathy, and know right from wrong. AI has no frontal lobe. It generates answers from patterns but cannot bear the human consequences of those answers. Responsibility must always rest with an accountable, identifiable person standing between the machine and any decision that touches someone’s life, liberty, or property.

Keep reading