May 5, 2020 – A new report from the Australian Strategic Policy Institute and the International Cyber Policy Centre detailed the state of rapidly developing “deep fake” technology and its potential to produce propaganda and misleading imagery more easily than ever.
The report, by Australian National University’s Senior Advisor for Public Policy Katherine Mansted and Researcher Hannah Smith, explained the costs of artificial intelligence technology allowing users to falsify or misrepresent existing media, as well as to generate new media entirely.
While audio-visual “cheap fakes” (edited media using tools other than AI) are not a recent phenomenon, the rapid rise of artificial-intelligence-powered technology has seen several means by which nefarious actors can produce misleading material at a staggering pace, four of which were highlighted by the ASPI report.
First, the face swapping method maps the face of one person and superimposes it onto the head of another.
The re-enactment method allows a deep fake creator to use facial tracking to manipulate the facial movements of their desired target. Another method, known as lip-syncing, combines re-enactment with phony audio generation to make it appear as though speakers are saying things they never did.
Finally, motion transfer technology allows the body movements of one person to control those of another.
This technology creates disastrous possibilities, the report said. When using various deep fake methods in conjunction, one can make it appear as though critical political figures are performing offensive or criminal acts or announcing forthcoming military action in hostile countries.
If deployed in a high-pressure situation where the prompt authentication of such media is not possible, real-life retaliation could occur.
The technology has already caused harm outside of the political arena.
The vast majority of deep fake technology is used on internet forums like Reddit to superimpose the faces of non-consenting peoples such as celebrities onto the bodies of men and women in pornographic videos, the report said.
Visual deep fakes are not perfect, and those available to the layman are often recognizable. But the technology has developed rapidly since 2017, and programs that work to make the deep fakes undetectable have as well.
Generative adversarial networks compete with other AI networks to develop and detect deep fakes, checking and refining hundreds or thousands of times, until deep fake audio and visual media are unrecognizable to the network and far less to the human eye. “GAN models are now widely accessible,” the report said, “and many are available for free online.”
Such forged videos are already widespread and may already have had an impact on public trust in elected officials and others, although such a phenomenon is difficult to quantify.
The report also detailed multiple instances in which a purposely altered video circulated online and potentially misinformed viewers, including a cheap fake video that was slowed and pitch-corrected to make House Speaker Nancy Pelosi appear inebriated.
Another video mentioned in the report, generated by AI thinktank Future Advocacy during the 2019 UK general election, used voice generation and lip-sync to make it appear as though now-Prime Minister Boris Johnson and then-opponent Jeremy Corbin were endorsing each other for the office.
Such videos can have a devastating effect on public trust, wrote Mansted and Smith. And in addition to the fact that the production of such videos is more accessible than ever, deep fake creators can use bots to swarm public internet forums and comment sections with commentary that, because of the lack of a visual element, can be almost impossible to recognize as artificial.
The accelerated production of such materials can make it feel as though the future of media is one where almost no video can be trusted to be authentic, and the report admitted that “On balance, detectors are losing the ‘arms race’ with creators of sophisticated deep fakes.”
However, Mansted and Smith concluded with several suggestions for combating the rise of ill-intentioned deep fakes.
Firstly, the report proposed that international governments and online forums should “fund research into the further development and deployment of detection technologies” as well as “require digital platforms to deploy detection tools, especially to identify and label content generated through deep fake processes.”
Secondly, the report suggested that media and individuals should stop accepting audio-visual media at face value, adding that “Public awareness campaigns… will be needed to encourage users to critically engage with online content.”
Such a change of perception will be difficult, however, as the spread of this imagery is largely based on emotion and not critical thinking.
Lastly, the report suggested the implementation of authentication standards such as encryption and blockchain technology.
“An alternative to detecting all false content is to signal the authenticity of all legitimate content,” Mansted and Smith wrote. “Over time, it’s likely that certification systems for digital content will become more sophisticated, in part mitigating the risk of weaponised deep fakes.”
Int’l Ethical Framework for Auto Drones Needed Before Widescale Implementation
Observers say the risks inherent in letting autonomous drones roam requires an ethical framework.
July 19, 2021 — Autonomous drones could potentially serve as a replacement for military dogs in future warfare, said GeoTech Center Director David Bray during a panel discussion hosted by the Atlantic Council last month, but ethical concerns have observers clamoring for a framework for their use.
Military dogs, trained to assist soldiers on the battlefield, are currently a great asset to the military. AI-enabled autonomous systems, such as drones, are developing capabilities that would allow them to assist in the same way — for example, inspecting inaccessible areas and detecting fires and leaks early to minimize the chance of on-the-job injuries.
However, concerns have been raised about the ability to impact human lives, including the recent issue of an autonomous drone possibly hunting down humans in asymmetric warfare and anti-terrorist operations.
As artificial intelligence continues to develop at a rapid rate, society must determine what, if any, limitations should be implemented on a global scale. “If nobody starts raising the questions now, then it’s something that will be a missed opportunity,” Bray said.
Sally Grant, vice president at Lucd AI, agreed with Bray’s concerns, pointing out the controversies surrounding the uncharted territory of autonomous drones. Panelists proposed the possibility of an international limitation agreement with regards to AI-enabled autonomous systems that can exercise lethal force.
Timothy Clement-Jones, who was a member of the U.K. Parliament’s committee on artificial intelligence, called for international ethical guidelines, saying, “I want to see a development of an ethical risk-based approach to AI development and application.”
Many panelists emphasized the immense risk involve if this technology gets in the wrong hands. Panelists provided examples stretching from terrorist groups to the paparazzi, and the power they could possess with that much access.
Training is vital, Grant said, and soldiers need to feel comfortable with this machinery while not becoming over-reliant. The idea of implementing AI-enabled autonomous systems into missions, including during national disasters, is that soldiers can use it as guidance to make the most informed decisions.
“AI needs to be our servant not our master,” Clement agreed, emphasizing that soldiers can use it as a tool to help them and not as guidance to follow. He compared AI technology with the use of phone navigation, pointing to the importance of keeping a map in the glove compartment in case the technology fails.
The panelists emphasized the importance of remaining transparent and developing an international agreement with an ethical risk-based approach to AI development and application in these technologies, especially if they might enter the battlefield as a reliable companion someday.
Deepfakes Could Pose A Threat to National Security, But Experts Are Split On How To Handle It
Experts disagree on the right response to video manipulation — is more tech or a societal shift the right solution?
June 3, 2021—The emerging and growing phenomenon of video manipulation known as deepfakes could pose a threat to the country’s national security, policy makers and technology experts said at an online conference Wednesday, but how best to address them divided the panel.
A deepfake is a highly technical method of generating synthetic media in which a person’s likeness is inserted into a photograph or video in such a way that creates the illusion that they were actually there. A well done deepfake can make a person appear to do things that they never actually did and say things that they never actually said.
“The way the technology has evolved, it is literally impossible for a human to actually detect that something is a deepfake,” said Ashish Jaiman, the director of technology operations at Microsoft, at an online event hosted by the Information Technology and Innovation Foundation.
Experts are wary of the associated implications of this technology being increasingly offered to the general population, but how best to address the brewing dilemma has them split. Some believe better technology aimed at detecting deepfakes is the answer, while others say that a shift in social perspective is necessary. Others argue that such a societal shift would be dangerous, and that the solution actually lies in the hands of journalists.
Deepfakes pose a threat to democracy
Such technology posed no problem when only Hollywood had the means to portray such impressive special effects, says Rep. Anthony Gonzalez, R-Ohio, but the technology has progressed to a point that allows most anybody to get their hands on it. He says that with the spread of disinformation, and the challenges that poses to establishing a well-informed public, deepfakes could be weaponized to spread lies and affect elections.
As of yet, however, no evidence exists that deepfakes have been used for this purpose, according to Daniel Kimmage, the acting coordinator for the Global Engagement Center of the Department of State. But he, along with the other panelists, agree that the technology could be used to influence elections and further already growing seeds of mistrust in the information media. They believe that its best to act preemptively and solve the problem before it becomes a crisis.
“Once people realize they can’t trust the images and videos they’re seeing, not only will they not believe the lies, they aren’t going to believe the truth,” said Dana Rao, executive vice president of software company Adobe.
New technology as a solution
Jaiman says Microsoft has been developing sophisticated technologies aimed at detecting deepfakes for over two years now. Deborah Johnson, emeritus technology professor at the University of Virginia School of Engineering, refers to this method as an “arms race,” in which we must develop technology that detects deepfakes at a faster rate than the deepfake technology progresses.
But Jaiman was the first to admit that, despite Microsoft’s hard work, detecting deepfakes remains a grueling challenge. Apparently, it’s much harder to detect a deepfake than it is to create one, he said. He believes that a societal response is necessary, and that technology will be inherently insufficient to address the problem.
Societal shift as a solution
Jaiman argues that people need to be skeptical consumers of information. He believes that until the technology catches up and deepfakes can more easily be detected and misinformation can easily be snuffed, people need to approach online information with the perspective that they could easily be deceived.
But critics believe this approach of encouraging skepticism could be problematic. Gabriela Ivens, the head of open source research at Human Rights Watch, says that “it becomes very problematic if people’s first reactions are not to believe anything.” Ivens’ job revolves around researching and exposing human rights violations, but says that the growing mistrust of media outlets will make it harder for her to gain the necessary public support.
She believes that a “zero-trust society” must be resisted.
Vint Cerf, the vice president and chief internet evangelist at Google, says that it is up to journalists to prevent the growing spread of distrust. He accused journalists not of deliberately lying, but often times misleading the public. He believes that the true risk of deepfakes lies in their ability to corrode America’s trust in truth, and that it is up to journalists to restore that trust already beginning to corrode by being completely transparent and honest in their reporting.
Complexity, Lack of Expertise Could Hamper Economic Benefits Of Artificial Intelligence
Artificial intelligence is said to open up a new age of economic development, but its complexity could hamper its rollout.
May 24, 2021 — One of the great challenges to adopting artificial intelligence is the lack of understanding of it, according to a panel hosted by the Atlantic Council’s new GeoTech Center.
The panel last week discussed the economic benefits of AI and how global policy leaders can leverage it to achieve sustainable economic growth with government buy-in. But getting the government excited and getting them to actually do something about it are two completely different tasks.
That’s because there exists little government understanding or planning around this emerging market, according to Keith Strier, vice-president of worldwide AI initiatives at NVIDIA, a tech company that designs graphics processing units.
If the trend continues, the consequences could be globally impactful, leading to a widening of the global economic divide and could even pose national security threats, he said.
“AI is the new critical infrastructure… It’s about the future of GDP,” said Strier.
Lack of understanding stems from complexity
The reason for a lack of government understanding stems from the complexity of AI research, and the lack of consensus among experts, Strier said. He noted that the metrics used to quantify AI performance are “deceptively complex” and technical. Experts struggle to even find consensus on defining AI, only adding to its already intrinsic complexity.
This divergence in expert opinion makes the research markedly difficult to break down and communicate to policy makers in digestible, useful ways.
“Policy is just not evidence based,” Strier said. “It’s not well informed.”
World economic divide could widen
Charles Jennings, AI entrepreneur and founder of internet technology company NeuralEye, warned of AI’s potential to widen the economic divide worldwide.
Currently, the 500 fastest computers in the world are split up between just 29 different countries, leaving the remaining 170 struggling to produce computing power. As computers become faster, the countries best suited to reap the economic benefits will do so at a rate that far outpaces less developed countries.
Jennings also believes that there exists security issues associated with the lack of AI understanding in government, claiming that the public’s increasing dependence on it matched with a lack of regulation could lead to a public safety threat. He is adamant that it’s time to bridge the gap between enterprise and policy.
Strier says there are three essential questions governments must answer: How much domestic AI compute capacity do we have? How does this compare to other nations? Do we have enough capacity to support our national AI ambitions?
Answering these questions would help governments address the AI question in terms of their own national values and interests. This would help create a framework that could mitigate the potential negative consequences which might otherwise affect us.
- Federal Trade Commission Will Likely Not Be Able to Implement Competition Rules, Panelists Say
- House Passes Ban on Chinese Equipment, 3.45 GHz Auction Reaches Reserve Price, Against a ‘Wi-Fi Tax’
- LEO Satellite Technology Should Be in All Schools, Gigabit Libraries Network Says
- Housing, Public Interest Groups Oppose Multitenant Exclusivity Agreements
- Broadband Breakfast on October 27, 2021 — When ‘Greenfield’ Fiber Meets ‘Brownfield’ Multiple Dwelling Units
- Federal Communications Commission Dispenses $544 Million in Rural Broadband Funds
Signup for Broadband Breakfast
Antitrust4 months ago
Experts Disagree Over Need, Feasibility of Global Standards for Antitrust Rules
Broadband Roundup2 months ago
Senators Intro App Bill, Groups Drop TracFone Buy Complaint, States Want Shorter Robocall Deadline
Infrastructure3 months ago
Lumen Responds to Allegations it Underbuilds While Collecting Public Funds
Broadband Roundup2 months ago
Mapping Comment Deadline Extended, AT&T Gets Federal Contract, 5G and LTE Drive Microwave Demand
Antitrust3 months ago
Daniel Hanley: Federal Communications Commission Must Block Verizon’s Acquisition of TracFone
#broadbandlive2 months ago
Broadband Breakfast on September 1, 2021 — What’s Next for Broadband Infrastructure Legislation?
Section 2303 months ago
Facebook, Google, Twitter Register to Lobby Congress on Section 230
Broadband Roundup2 months ago
FCC and FTC Announce Open Meeting Agendas and AT&T Signs Deal with OneWeb