Facebook, Google activate SOS features after Barcelona attack

Washington, Jan 7: Facebook will ban deepfake videos ahead of the US elections but the new policy will still allow heavily edited clips so long as they are parody or satire, the social media giant said Tuesday.

Deepfake videos are hyper-realistic doctored clips made using artificial intelligence or programs that have been designed to accurately fake real human movements.

In a blog published following a Washington Post report, Facebook said it would begin removing clips that were edited--beyond for clarity and quality--in ways that "aren't apparent to an average person" and could mislead people.

Clips would be removed if they were "the product of artificial intelligence or machine learning that merges, replaces or superimposes content onto a video, making it appear to be authentic," the statement from Facebook vice-president Monika Bickert said.

However, the statement added: "This policy does not extend to content that is parody or satire, or video that has been edited solely to omit or change the order of words."

US media noted the new guidelines would not cover videos such as the 2019 viral clip -- which was not a deepfake -- of House Speaker Nancy Pelosi that appeared to show her slurring her words.

Facebook also gave no indication on the number of people assigned to identify and take down the offending videos, but said videos failing to meet its usual guidelines would be removed, and those flagged clips would be reviewed by teams of third-party fact-checkers -- among them AFP.

The news agency has been paid by the social media giant to fact-check posts across 30 countries and 10 languages as part of a program starting in December 2016, and including more than 60 organisations.

Content labeled "false" is not always removed from newsfeeds but is downgraded so fewer people see it -- alongside a warning explaining why the post is misleading.

Los Angeles, Apr 28: People who experience loss of smell as one of the COVID-19 symptoms are likely to have a mild to moderate clinical course of the disease, according to a study which may help health care providers determine which patients require hospitalisation.

The findings, published in the journal International Forum of Allergy & Rhinology, follows an earlier study that validated the loss of smell and taste as indicators of infection with the novel coronavirus, SARS-CoV-2.

According to the scientists from the University of California (UC) San Diego Health in the US, patients who reported loss of smell were 10 times less likely to be hospitalised for COVID-19 compared to those without the symptom.

"One of the immediate challenges for health care providers is to determine how to best treat persons infected by the novel coronavirus," said Carol Yan, first author of the current study and rhinologist from the UC San Diego Health.

"If they display no or mild symptoms, can they return home to self-quarantine or will they likely require hospitalisation? These are crucial questions for hospitals trying to efficiently and effectively allocate finite medical resources," Yan said.

The findings, according to the researchers, suggest that loss of smell may be predictive of a milder clinical course of COVID-19.

"What's notable in the new findings is that it appears that loss of smell may be a predictor that a SARS-CoV-2 infection will not be as severe, and less likely to require hospitalisation," Yan said.

"If an infected person loses that sense, it seems more likely they will experience milder symptoms, barring other underlying risk factors," she added.

Risk factors for COVID-19 previously reported by other studies include age, and underlying medical conditions, such as chronic lung disease, serious heart conditions, diabetes, and obesity.

In the current study, the scientists made a retrospective analysis between March 3 and April 8 including 169 patients who tested positive for COVID-19 at UC San Diego Health.

They assessed olfactory and gustatory data for 128 of the 169 patients, 26 of whom required hospitalisation.

According to the researchers, patients who were hospitalised for COVID-19 treatment were significantly less likely to report anosmia or loss of smell -- 26.9 per cent compared to 66.7 per cent for COVID-19-infected persons treated as outpatients.

Similar percentages were found for loss of taste, known as dysgeusia, they said.

"Patients who reported loss of smell were 10 times less likely to be admitted for COVID-19 compared to those without loss of smell," said study co-author Adam S. DeConde.

"Moreover, anosmia was not associated with any other measures typically related to the decision to admit, suggesting that it's truly an independent factor and may serve as a marker for milder manifestations of Covid-19," DeConde said.

The researchers suspect that the findings hint at some of the physiological characteristics of the infection.

"The site and dosage of the initial viral burden, along with the effectiveness of the host immune response, are all potentially important variables in determining the spread of the virus within a person and, ultimately, the clinical course of the infection," DeConde said.

If the SARS-CoV-2 virus initially concentrates in the nose and upper airway, where it impacts olfactory function, that may result in an infection that is less severe and sudden in onset, decreasing the risk of overwhelming the host immune response, respiratory failure, and hospitalisation, the scientists added.

"This is a hypothesis, but it's also similar to the concept underlying live vaccinations," DeConde explained.

"At low dosage and at a distant site of inoculation, the host can generate an immune response without severe infection," he added.

Loss of smell, according to the study, might also indicate a robust immune response which has been localised to the nasal passages, limiting effects elsewhere in the body.

Citing the limitations of the study, the scientists said they relied upon self-reporting of anosmia from participants, which posed a greater chance of recall bias among patients once they had been diagnosed with COVID-19.

They added that patients with more severe respiratory disease requiring hospitalisation may not be as likely to recognise or recall the loss of smell.

So the researchers said more expansive studies are needed for validating the results.

Feb 27: With the window to submit comments on India's proposed personal data protection law closing on Tuesday, a period of anxious wait for final version of the Bill started for social media firms.

This comes even as global Internet companies have called on the government for improved transparency related to intermediary Guidelines (Amendment) Rules and allay fears about the prospect of increased surveillance and prompting a fragmentation of the Internet in India that would harm users.

As per the proposed amendments, an intermediary having over 50 lakh users in the country will have to be incorporated in India with a permanent registered office and address.

When required by lawful order, the intermediary shall, within 72 hours of communication, provide such information or assistance as asked for by any government agency or assistance concerning security of the state or cybersecurity.

This means that the government could pull down information provided by platforms such as Wikipedia, potentially hampering its functioning in India.

In the open letter to IT Minister Ravi Shankar Prasad, leading browser and software development platform like Mozilla, Microsoft-owned GitHub and Cloudflare earlier called for improved transparency by allowing the public an opportunity to see a final version of these amendments prior to their enactment.

According to a Business Insider report, Indian users may lose access to Wikipedia if the new intermediary rules for internet and social media companies are approved.

Since the rules would require the website to take down content deemed illegal by the government, it would require Wikipedia to show different content for different countries.

Anusha Alikhan, senior communications director for Wikimedia told Business Insider that the platform is built though languages and not geographies. Therefore, removing content from one country, while it is still visible to other country users may not work for the company’s model.

India is one of Wikipedia’s largest markets. Over 771 million Indian users accessed the site in just November 2019.

Also read: Explained: What is the Personal Data Protection Bill and why you should care

The Personal Data Protection Bill, 2019, which was introduced in Lok Sabha in the winter session last year, was referred to a Joint Parliamentary Committee (JPC) of both the Houses.

The government last month decided to seek views and suggestions on the Bill from individuals and associations and bodies concerned and the last date for submitting the comments was on Tuesday.

Prasad, while introducing the Personal Data Protection Bill, 2019, in the Lok Sabha on December 11, announced that the draft Bill empowers the government to ask companies including Facebook, Google and others for anonymised personal data and non-personal data.

There was a buzz when the Bill's latest version was introduced in the Lok Sabha, especially the provision seeking to allow the use of personal and non-personal data of users in some cases, especially when national security is involved.

Several legal experts red-flagged the issue and said the provision will give the government unaccounted access to personal data of users in the country.

In their submission to the JPC, several organisations also flagged that the power to collect non-personal and anonymised data by the government without notice and consent should not form part of the Bill because of issues regarding effective anonymisation and potential abuse.

"Clauses 35 and 36 of the Bill provide unbridled access to personal data to the Central Government by giving it powers to exempt its agencies from the application of the Bill on the basis of various broad worded grounds," SFLC.in, a New Delhi-based not-for-profit legal services organisation, commented.

The Software Alliance, also known as BSA, a trade group which includes tech giants such as Microsoft, IBM and Adobe, among others said that the current version of the privacy bill pose substantial challenges, including the sweeping new powers for the government to acquire non-personal data, restrictions on data transfers, and local storage requirements.

"We urge the Joint Parliamentary Committee, as it considers revisions to the Bill, to eliminate provisions concerning non-personal data from the Personal Data Protection Bill and to remove the data localisation requirements and restrictions on international data flows," said Venkatesh Krishnamoorthy, Country Manager-India, BSA.

The Personal Data Protection (PDP) Bill, 2019 draws its origins from the Justice B.N. Srikrishna Committee on data privacy, which produced a draft of legislation that was made public in 2018 ("the Srikrishna Bill").

The mandatory requirement for storing a mirror copy of all personal data in India as per Section 40 of the Srikrishna Bill has been done away with in the PDP Bill, 2019, meaning that companies like Facebook and Twitter would be able to store data of Indian users abroad if they so wish.

But the bill prohibits processing of sensitive personal data and critical personal data outside India.

What is more, what constitutes critical data has not been clearly defined.

As per the proposals, social media companies will have to modify their application as they are required to have a system in place by which a user can verify themselves.

So legal experts believe that some system to upload identification documents should be there and something like the Twitter blue tick mark should be there to identify verified accounts.

"The 2019 Bill introduces a new category of data fiduciaries called social media intermediaries ('SMIs'). SMIs are a subcategory of significant data fiduciaries ('SDFs') and will be notified by the Central government after due consultation with the DPA, or the Data Protection Authority. Clause 26(4) of the Bill defines SMIs as intermediaries who primarily or solely enable online interaction between two or more users," SFLC.in said.

"On a plain reading of the definition, online platforms like Facebook, Twitter, YouTube, TikTok, ShareChat and WhatsApp are likely to be notified as SMIs under the Bill," it added.