An independent expert group tasked with advising the European Commission to inform its regulatory response to artificial intelligence -- to underpin EU lawmakers' stated aim of ensuring AI developments are "human centric" -- has published its policy and investment recommendations. This follows earlier ethics guidelines for "trustworthy AI", put out by the High Level Expert Group (HLEG) for AI back in April, when the Commission also called for participants to test the draft rules. The AI HLEG's full policy recommendations comprise a highly detailed 50-page document -- which can be downloaded from this web page. The group, which was set up in June 2018, is made up of a mix of industry AI experts, civic society representatives, political advisers and policy wonks, academics and legal experts. The document includes warnings on the use of AI for mass surveillance and scoring of EU citizens, such as China's social credit system, with the group calling for an outright ban on "AI-enabled mass scale scoring of individuals".

ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up for ProPublica's Big Story newsletter to receive stories like this one in your inbox as soon as they are published. Ariella Russcol specializes in drama at the Frank Sinatra School of the Arts in Queens, New York, and the senior's performance on this April afternoon didn't disappoint. While the library is normally the quietest room in the school, her ear-piercing screams sounded more like a horror movie than study hall. But they weren't enough to set off a small microphone in the ceiling that was supposed to detect aggression.

An organization that won the Nobel Prize in 2017 for its work to eliminate nuclear weapons is sounding the alarm about the possibility of artificial intelligence leading to unintended wars. Beatrice Fihn, executive director of the International Campaign to Abolish Nuclear Weapons, is worried that hackers could breach A.I. technologies that are used in nuclear programs or that they could use A.I. to dupe countries into launching attacks. For example, deepfakes, or realistic-looking computer-altered videos, may be used to "create a perceived threat that might not be there," she warns, prompting governments to overreact. Fihn told Fortune that she wants to convene a meeting in the fall with nuclear weapons experts and some of the leading companies in A.I. and cybersecurity. Participants in the off-the-record event, she said, would produce a document that her group would use to inform governments and others about the danger.

The Allen Institute for Artificial Intelligence has an interesting new tactic in the war on fake news: make more of it. A team of researchers at the institute recently developed Grover, a neural network capable of generating fake news articles in the style of actual human journalists. In essence, the group is fighting fire with fire because the better Grover gets at generating fakes, the better it'll be at detecting them. Our study presents a surprising result: the best way to detect neural fake news is to use a model that is also a generator. The generator is most familiar with its own habits, quirks, and traits, as well as those from similar AI models, especially those trained on similar data, i.e. publicly available news.

AI is transforming the health industry. As a society we are becoming more data hungry than ever before, and this is also evident at an individual level though our growing fascination with wearable technology and e-health. Previously, we've summed up AI and deep learning from a beginner's perspective, and discussed some of their use cases in healthcare, specifically medical diagnoses. Below, we continue in healthcare, providing a brief overview of deep learning in drug discovery, e-health and electronic health records. Drug discovery from idea conception to a marketable product, can take over decade, and on average costs US$2.6 billion.

We all know the old saying: garbage in, garbage out. This has been especially true with early trails of artificial intelligence. Humans building the algorithms are inherently flawed and have deeply ingrained biases in their thought processes, and this translates to bias in the output of many artificial intelligence algorithms. We've seen one algorithm learn that male job candidates are preferred to female job candidates and automatically kick out not only the resumes of women, but also those that listed women as references. Building ethical AI is tricky, but it can, and must, be done.

It may be telling that one of the smash hit books of the past year has been on sleep. Berkeley professor of neuroscience Matthew Walker's Why We Sleep was a comprehensive analysis of why humans need to sleep, which doubled up as a polemic against the habits that lead to insomnia. He marvels: "A hundred years ago, less than 2 per cent of the population in the United States slept six hours or less a night. Now, almost 30 per cent of American adults do." The health implications are shocking. Lack of sleep leads to mental illness, diabetes and chaos.

Military doctrine identifies five domains of warfare--land, sea, air, space and information. While borders and barriers define the four natural domains, the fifth dimension, with the advancements of artificial intelligence, is rapidly expanding with the potential to destabilize free and open international order. Nations like China and Russia are making significant investments in AI for military purposes, potentially threatening world norms and human rights. This year the Defense Department, in support of the National Defense Strategy, launched its Artificial Intelligence Strategy in concert with the White House executive order creating the American Artificial Intelligence Strategy. The DoD AI strategy states the U.S., together with its allies and partners, must adopt AI to maintain its strategic position, prevail on future battlefields and safeguard order.

GANpaint Studio could also be used to improve and debug other GANs that are being developed, by analyzing them for "artifact" units that need to be removed. In a world where opaque AI tools have made image manipulation easier than ever, it could help researchers better understand neural networks and their underlying structures. "Right now, machine learning systems are these black boxes that we don't always know how to improve, kind of like those old TV sets that you have to fix by hitting them on the side," says Bau, lead author on a related paper about the system with a team overseen by Torralba. "This research suggests that, while it might be scary to open up the TV and take a look at all the wires, there's going to be a lot of meaningful information in there." One unexpected discovery is that the system actually seems to have learned some simple rules about the relationships between objects.

San Francisco supervisors approved a ban on police using facial recognition technology, making it the first city in the U.S. with such a restriction. LOS ANGELES -- A major supplier of body cameras to law enforcement agencies across the country has decided to forgo selling facial recognition technology with its products. Axon, which supplies 48 police departments in major cities with body cameras, made the decision after the company's ethics board concluded the technology was not accurate enough to be implemented in the field and could potentially cause major trust issues between law enforcement and their communities. In a 42-page report, the ethics board detailed concerns with the inaccuracy of the software, saying results showed it was less accurate when identifying women, younger people, and "worsens when trying to identify people of color compared to white people, a troubling disparity that would only perpetuate or exacerbate the racial inequities that cut across the criminal justice system." Is Facebook listening to me?