Your newsroom needs an AI ethics policy. Start here.; Poynter, March 25, 2024

Kelly McBride , Poynter; Your newsroom needs an AI ethics policy. Start here.

"Every single newsroom needs to adopt an ethics policy to guide the use of generative artificial intelligence. Why? Because the only way to create ethical standards in an unlicensed profession is to do it shop by shop.

Until we create those standards — even though it’s early in the game — we are holding back innovation.

So here’s a starter kit, created by Poynter’s Alex Mahadevan, Tony Elkins and me. It’s a statement of journalism values that roots AI experimentation in the principles of accuracy, transparency and audience trust, followed by a set of specific guidelines.

Think of it like a meal prep kit. Most of the work is done, but you still have to roll up your sleeves and do a bit of labor. This policy includes blank spaces, where newsroom leaders will have to add details, saying “yes” or “no” to very specific activities, like using AI-generated illustrations.

In order to effectively use this AI ethics policy, newsrooms will need to create an AI committee and designate an editor or senior journalist to lead the ongoing effort. This step is critical because the technology is going to evolve, the tools are going to multiply and the policy will not keep up unless it is routinely revised."

He Hunts Sloppy Scientists. He’s Finding Lots of Prey.; The New York Times, February 2, 2024

Matt Richtel, The New York Times ; He Hunts Sloppy Scientists. He’s Finding Lots of Prey.

"Sholto David, 32, has a Ph.D. in cellular and molecular biology from Newcastle University in England. He is also developing an expertise in spotting errors in scientific papers. Most recently, and notably, he discovered flawed or manipulated data in studies conducted by top executives at the Harvard-affiliated Dana-Farber Cancer Institute. The institute said that it was requesting retraction of six manuscripts and had found 31 other manuscripts that required corrections.

From his home in Wales, Dr. David scours new research publications for images that are mislabeled and manipulated, and he regularly finds mistakes, or malfeasance, in some of the most prominent scientific journals. Accuracy is vital, as peer-reviewed papers often provide the evidence for drug trials or further lines of research. Dr. David said that the frequency of such errors suggests an underlying problem for science.

His interview with The New York Times has been edited and condensed...

Does this call into question the peer-review process?

I think that’s something that people need to think about. These are top scientific journals with errors that escaped peer review. Maybe the peer reviewers are looking for other things. Maybe they like to look at the methods or the conclusions more carefully than the results. But, yeah, it does make me think that people should question how effective the peer-review process has been."

After Hospital Blast, Headlines Shift With Changing Claims; The New York Times, October 18, 2023

 Katie Robertson, The New York Times; After Hospital Blast, Headlines Shift With Changing Claims

"Covering wars is always fraught, both because journalists on the ground are often in harm’s way and because the sides at war aggressively push information in their favor."

Senator wants Google to answer for accuracy, ethics of generative AI tool; HealthcareITNews, August 9, 2023

 Mike Miliard, HealthcareITNews; Senator wants Google to answer for accuracy, ethics of generative AI tool

"Sen. Mark Warner, D-Virginia, wrote a letter to Sundar Pichai, CEO of Google parent company Alphabet, on Aug. 8, seeking clarity into the technology developer's Med-PaLM 2, an artificial intelligence chatbot, and how it's being deployed and trained in healthcare settings."

Managing the Risks of Generative AI; Harvard Business Review (HBR), June 6, 2023

Kathy Baxter 

and 

Yoav Schlesinger

, Harvard Business Review (HBR); Managing the Risks of Generative AI

"Guidelines for the ethical development of generative AI

Our new set of guidelines can help organizations evaluate generative AI’s risks and considerations as these tools gain mainstream adoption. They cover five focus areas."

About WBUR's Ethics Guide; WBUR, March 10, 2022

WBUR; About WBUR's Ethics Guide

"The committee approached the guidelines from the vantage point of WBUR journalists and journalism — while acknowledging the importance of the ethical guidelines and standards that need to be understood and embraced by everyone who works or is associated with WBUR.

The committee used the NPR Ethics Handbook as a structural model and source text, adopted with a WBUR voice. They also addressed ethics issues from a 2021 perspective, recognizing that much has changed in the public media and journalism field since the NPR Handbook was first written a decade ago."

WBUR Ethics Guide PDF: https://d279m997dpfwgl.cloudfront.net/wp/2022/03/WBUR-Ethics-Guidelines.pdf  

Wrongfully Accused by an Algorithm; The New York Times, June 24, 2020

Kashmir Hill, The New York Times; Wrongfully Accused by an Algorithm

In what may be the first known case of its kind, a faulty facial recognition match led to a Michigan man’s arrest for a crime he did not commit.

"Clare Garvie, a lawyer at Georgetown University’s Center on Privacy and Technology, has written about problems with the government’s use of facial recognition. She argues that low-quality search images — such as a still image from a grainy surveillance video — should be banned, and that the systems currently in use should be tested rigorously for accuracy and bias.

“There are mediocre algorithms and there are good ones, and law enforcement should only buy the good ones,” Ms. Garvie said.

About Mr. Williams’s experience in Michigan, she added: “I strongly suspect this is not the first case to misidentify someone to arrest them for a crime they didn’t commit. This is just the first time we know about it.”"

Consumer DNA Testing May Be the Biggest Health Scam of the Decade; Gizmodo, November 20, 2019

Ed Cara, Gizmodo; Consumer DNA Testing May Be the Biggest Health Scam of the Decade

"This test, as well as many of those offered by the hundreds of big and small DNA testing companies on the market, illustrates the uncertainty of personalized consumer genetics.

The bet that companies like 23andMe are making is that they can untangle this mess and translate their results back to people in a way that won’t cross the line into deceptive marketing while still convincing their customers they truly matter. Other companies have teamed up with outside labs and doctors to look over customers’ genes and have hired genetic counselors to go over their results, which might place them on safer legal and medical ground. But it still raises the question of whether people will benefit from the information they get. And because our knowledge of the relationship between genes and health is constantly changing, it’s very much possible the DNA test you take in 2020 will tell you a totally different story by 2030."

Real-Time Surveillance Will Test the British Tolerance for Cameras; The New York Times, September 15, 2019

Adam Satariano, The New York Times; Real-Time Surveillance Will Test the British Tolerance for Cameras

Facial recognition technology is drawing scrutiny in a country more accustomed to surveillance than any other Western democracy. 

"“Technology is driving forward, and legislation and regulation follows ever so slowly behind,” said Tony Porter, Britain’s surveillance camera commissioner, who oversees compliance with the country’s surveillance camera code of practice. “It would be wrong for me to suggest the balance is right.”

Britain’s experience mirrors debates about the technology in the United States and elsewhere in Europe. Critics say the technology is an intrusion of privacy, akin to constant identification checks of an unsuspecting public, and has questionable accuracy, particularly at identifying people who aren’t white men."

Parkland school turns to experimental surveillance software that can flag students as threats; The Washington Post, February 13, 2019

Drew Harwell, The Washington Post; Parkland school turns to experimental surveillance software that can flag students as threats

"The specter of student violence is pushing school leaders across the country to turn their campuses into surveillance testing grounds on the hope it’ll help them detect dangerous people they’d otherwise miss. The supporters and designers of Avigilon, the AI service bought for $1 billion last year by tech giant Motorola Solutions, say its security algorithms could spot risky behavior with superhuman speed and precision, potentially preventing another attack.

But the advanced monitoring technologies ensure that the daily lives of American schoolchildren are subjected to close scrutiny from systems that will automatically flag certain students as suspicious, potentially spurring a response from security or police forces, based on the work of algorithms that are hidden from public view.

The camera software has no proven track record for preventing school violence, some technology and civil liberties experts argue. And the testing of their algorithms for bias and accuracy — how confident the systems are in identifying possible threats — has largely been conducted by the companies themselves."

White House shares doctored video to support punishment of journalist Jim Acosta; The Washington Post, November 8, 2018

Drew Harwell, The Washington Post; White House shares doctored video to support punishment of journalist Jim Acosta

"The video has quickly become a flashpoint in the battle over viral misinformation, turning a live interaction watched by thousands in real time into just another ideological tug-of-war. But it has also highlighted how video content — long seen as an unassailable verification tool for truth and confirmation — has become as vulnerable to political distortion as anything else."

Daniel Radcliffe and the Art of the Fact-Check; The New Yorker, October 15, 2018 Issue

Michael Schulman, The New Yorker; Daniel Radcliffe and the Art of the Fact-Check

 [Kip Currier: I saw Daniel Radcliffe on MSNBC's Morning Joe program today, talking about the debut of a new Broadway play, "The Lifespan of a Fact", in which he stars, along with Bobby Cannavale and Cherry Jones. The play's exploration of "truth" is timely and intriguing.

A bit later, I heard reporter Robert Costa on MSNBC's Stephanie Ruhle discussing "disappeared" and presumably murdered Saudi journalist Jamal Koshaggi's final piece for The Washington Post, which was submitted before he was last seen and was published yesterday, with a note from his editor: "Jamal Khashoggi: What the Arab world needs most is free expression". Koshaggi's observations are prescient, poignant, and thought-provoking. Costa made the excellent point that truth goes hand-in-hand with the ability to have the freedom to seek and report truth.]

"Fact: the actor Daniel Radcliffe is currently starring in the Broadway show “The Lifespan of a Fact,” as a magazine fact checker with an aviation inspector’s zeal for accuracy. The play is drawn from a real-life skirmish: in 2005, Jim Fingal, an intern at The Believer, was tasked with fact-checking an essay by John D’Agata (played by Bobby Cannavale), about a teen suicide in Las Vegas. D’Agata had more of a watercolorist’s approach to the truth. When Fingal tried to correct his claim that Las Vegas had thirty-four licensed strip clubs—a source indicated that it was thirty-one—D’Agata said that he liked the “rhythm” of thirty-four. Their epistolary tussle was expanded into a book in 2012.

Not long ago, Radcliffe arrived at the offices of this magazine, wearing a maroon cap and a green jacket and clutching a latte. He had come to try his own hand at fact-checking, with the help of The New Yorker’s fact-checking department."

Unchecked fake news gave rise to an evil empire in Star Wars; Washington Post, May 4, 2017

Ben Guarino, Washington Post; Unchecked fake news gave rise to an evil empire in Star Wars

[Kip Currier: In discussing the "media-poor" fictional Star Wars-universe, the Washington Post reporter and cited experts implicate the critical roles of "real world" archives, libraries, the historical record, print and media cultures, education, access to information, data stewardship and analysis, rational "truth"-based discourse, free and independent press, and literate and questioning citizenry, within technology-infused--and increasingly tech-dependent--societies. These implications also raise some persuasive arguments for the relevance and interdependence of humanities-focused practitioners and research disciplines to technology fields and endeavors.]

"“Fake news in 'Star Wars' is probably their number one problem,” says Ryan Britt, an editor who specializes in science fiction at the website Inverse. Britt, in his 2015 book “Luke Skywalker Can’t Read,” makes a provocative claim: Most “Star Wars” denizens, if they're not illiterate, seem fundamentally disinterested in reading...

Fake news is a deadly symptom of the media-poor culture displayed in “Star Wars.” Facebook, in a report released at the end of April, defined fake news as a “catch-all” phrase that may include “hoaxes, rumors, memes, online abuse, and factual misstatements by public figures that are reported in otherwise accurate news pieces.” And in “Star Wars,” a few whopping “factual misstatements” by a public figure give rise to an evil empire.

Near the end of the prequel “Revenge of the Sith,” the elected leader of the Galactic Republic gives a speech. It's a rousing speech, full of carnage and conspiracy. Supreme Chancellor Palpatine spins a wild theory that the powerful elite, the Jedi, wish to subvert the government. It's also total bull...

“When you take out print, when you legislate against media, what results is some kind of totalitarian state,” says Joseph Hurtgen, an English instructor at Georgia's Young Harris College and an expert in archival theory, the way information is kept and stored. “That’s always where this goes when you undermine print culture.”

The funny thing about records in “Star Wars,” Hurtgen says, is that they betray an obsession with technology. “The only archive that anybody bothers to keep in 'Star Wars' is technology,” he says. “Nobody’s writing down memos or news.”

Even those technological archives are devoid of context. The Jedi library contains volumes of star charts but allows no room for questioning their accuracy. “The library is complete garbage,” in Britt's estimation."

Oscars Mistake Casts Unwanted Spotlight on PwC; New York Times, February 27, 2017

David Gelles and Sapna Maheshwari, New York Times; 

Oscars Mistake Casts Unwanted Spotlight on PwC

"One video posted there, introducing Mr. Cullinan and Ms. Ruiz, began with the line, “The reason we were even first asked to take on this role was because of the reputation PwC has in the marketplace for being a firm of integrity, of accuracy and confidentiality.” It went on to note that the relationship was “symbolic of how we’re thought of beyond this role and how our clients think of us.”

But how clients think of PwC may change.

Mr. Gilman, the crisis communications specialist, said he was curious to see if PwC kept the Oscars contract. “They have branded themselves around this event saying, ‘We’re trusted’ — that’s the implication. Now I think that will take a hit.”"

Fact check: This is not really a post-fact election; Washington Post, 10/7/16

Alexios Mantzarlis, Washington Post; Fact check: This is not really a post-fact election:

"Unique visitors to The Washington Post’s Fact Checker were up 477 percent year-over-year in July, and up again in August. NPR recorded the highest traffic in the history of its website thanks to its live fact-checking of the first presidential debate. At least 6 million people had checked out the annotated transcript by the next morning. PolitiFact racked up 3.5 million page views in the 24 hours after that debate, drawing more traffic in one day than it did in entire months during the 2012 presidential campaign.

So voters want more fact-checking. But is it making any difference? Do people change their minds when faced with a fact check that surprises them, or do they internalize only fact checks that suit their own biases? Our understanding of basic psychology suggests that fact checks are often read with a partisan eye.

Often, but not always. A new working paper by Brendan Nyhan of Dartmouth College and Jason Reifler of Exeter University, who have studied fact-checking extensively, indicates that readers can learn from fact checks."

The 'p-word' problem: Trump's comments pose issue for news outlets; CNN, 10/7/16

[Graphic Language] Frank Pallotta, CNN; The 'p-word' problem: Trump's comments pose issue for news outlets:

"The 2005 videotape in which Donald Trump can be heard making vulgar comments about women posed a dilemma for news outlets: do they run in full the most vital and graphic line of a news story that could help determine a presidential election -- or do they censor it for the sake of decency?"

Improving the Trademark Register; Director's Forum: A Blog from USPTO's Leadership, 10/5/16

Guest Blog from Commissioner for Trademarks Mary Boney Denison, Director's Forum: A Blog from USPTO's Leadership; Improving the Trademark Register:

"When selecting a mark for a new product or service, a business will search the USPTO database of registered marks to determine whether a particular mark is available. Registered trademarks that are not actually in use in commerce unnecessarily block someone else from registering the mark.

To ensure the accuracy of our trademark registry, in 2012, the USPTO launched a pilot program to gather data on whether registered marks were actually being used on the products and services listed on their registrations.

During the pilot, in 500 randomly-selected maintenance filings we required the registrant to submit proof of use for two additional items for each class listed on the registration. Although the registrant must submit one example of use per class in a maintenance filing, typically the registration will list multiple products or services for each class.

At the conclusion of the pilot, the USPTO determined that in more than half of the trademark registrations selected, the owner was unable to verify the actual use of the mark for the goods or services queried. This was in spite of the owner having recently sworn under penalty of perjury to such ongoing use as part of the maintenance filing. We issued a report on the results and held a roundtable to discuss the results and next steps. The consensus among roundtable participants was that the results of the pilot program indicated a need for some action to improve the accuracy and integrity of the register. As a result of these findings and input from the trademark community, we are now taking a three-pronged approach to tackling the so-called “deadwood” in our searchable database of registered marks."

The F-word is making more frequent appearances in appellate opinions; ABA Journal, 8/29/16

Debra Cassens Weiss, ABA Journal; The F-word is making more frequent appearances in appellate opinions:

"Do you call it the F-word? Use asterisks instead of some of the letters? Substitute the word “expletive”? Or write it out in full?

Appellate opinions are more frequently choosing the last option, Law.com reports."

Someone is wrong on the Internet. That’s where I come in.; Washington Post, 8/19/16

Brooke Binkowski, Washington Post; Someone is wrong on the Internet. That’s where I come in. :

"...[T]his work is a social good. In addition to the hate mail, we receive a lot of frightened mail, from people who aren’t certain about their place in an increasingly scary world, where danger seemingly lurks around every corner. Times are changing, yes, but some websites take advantage of that uncertainty by plucking stories from the news and fluffing them up to make it appear as though there’s going to be a major disaster any minute now. (Recent examples included activity on a major fault line that supposedly signified an imminent massive earthquake on the West Coast, and the closure of cargo routes in the North Atlantic. Neither of these stories were true.) Concerned people pass along the stories, “just in case,” and spread the anxiety further. The people who own clickbait sites are never held accountable for dealing in fear. They do, however, make quite a lot of money from advertising.

We don’t think that our work will affect people committed to their belief systems to the exclusion of all facts. But Snopes can be a place where people begin their own research; we can be a reference for people who care to excavate the facts behind the often-terrifying headlines. We don’t pretend to be, nor do we want to be, the final word on any subject. We would like to be a starting point, though. In cases where clickability and virality trump fact, we feel that knowledge is the best antidote to fear."

Can mythbusters like Snopes.com keep up in a post-truth era?; Guardian, 8/1/16

Rory Carroll, Guardian; Can mythbusters like Snopes.com keep up in a post-truth era? :

"Mikkelson lists four principal misinformation sources:

1 Legitimate satire sites such as the Onion, which dupe the truly credulous, requiring occasional intervention. “No, SeaWorld isn’t drowning live elephants as part of a new attraction.” “Are the parents of teen Caitlin Teagart going to euthanise her because she is only capable of texting and rolling her eyes? False.”

2 Legitimate news organisations that regurgitate stories without checking, such as the $200 Bill Clinton haircut on Air Force One which supposedly snarled air traffic at LAX in 1993.

3 Political sites that distort, such as Breitbart.com twisting an Obama quote about the “contributions of Muslim Americans to building the very fabric of our nation” into the headline “Obama: Muslims Built ‘The Very Fabric of Our Nation’.”

4 Fake news sites fabricating click-bait stories. Such as: “Ted Cruz sent shockwaves through the Republican Party today when he announced he would endorse Donald Trump for President, but only if the GOP nominee would publicly support a ban on masturbation, (saying) without ‘swift action … the country was doomed to slide down a slippery slope of debauchery and self-satisfaction’.” Snopes sourced this to a site that mimicked ABC News to lure clicks to an underlying malware site, generating advertising revenue. It named and shamed the worst offenders earlier this year."