Ethics group files complaint to the FTC against OpenAI’s, Microsoft slips ads in new Bing bot and Cloudflare uses lava lamps to generate randomness for better encryption.
 |
 |
 |
These stories and more on Hashtag Trending for Friday, March 31st.
I’m your guest host for the week, James Roy – here’s today’s top tech news stories.
The AI arms race has taken the tech world by a storm and it’s not slowing down. But ethics groups and experts are now worried.
Yesterday, the Center for AI and Digital Policy (CAIDP) filed a complaint with the Federal Trade Commission (FTC) arguing that GPT-4, OpenAI’s latest juiced up large language model, violates the FTC’s rules against unfair and deceptive practices.
This move comes after AI experts and – Elon Musk signed an open letter urging AI labs around the world to pause the development of large-scale AI projects.
The high-profile letter published, early this week by nonprofit Future of Life Institute, notes that AI labs are currently in an “out-of-control race” to develop and deploy machine learning systems that “no one, not even their creators can understand, predict or reliably control”.
The signatories also cited fears over the “profound risks to society and humanity” they believe this technology could pose.
The CAIDP complaint detailed these very risks, such as GPT-4’s capacity to make up false information, produce malicious code, highly tailored propaganda or race or gender-based biases in things like hiring.
The ethics group also argued that GPT-4 crosses a line of consumer harm that should invite regulatory action from the FTC.
The complaint claims, “OpenAI released GPT-4 to the public for commercial use with full knowledge of these risks,” including potential bias and harmful behavior.
As part of the complaint, CAIDP is asking the FTC to halt any further commercial deployment of GPT models and require independent assessments of the models before they are rolled out.
The complaint also asked for a publicly accessible reporting tool similar to the one where consumers can file fraud complaints.
Source: The Verge
Sixty percent of organizations hit by an authentication-related cyberattack in the last 12 months say it could have been avoided with a passwordless system. Here’s why.
Organizations have been using two-factor authentication but that barely slowed attacks as gullible users often get duped into providing their credentials to bad actors.
Plus many still use legacy authentication like username and password, password manager or single sign-on.
Only a meager 28 per cent used some form of password-less authentication.
That, despite many organizations pointing out several pain points with legacy authentication methods. Issues raised included difficulty with securely authenticating remote workers, unmanaged third-party devices, technology complexity, employee resistance to adoption and password reset.
As a matter of fact, 81 per cent of enterprises admitted having trouble accessing work-critical information on occasions they forgot a password. Companies even report spending $375 per employee per year on password issues.
But the benefits of passwordless are increasingly apparent to organizations.
Improving user experience and productivity, strengthening cybersecurity, encouraging employee adoption of multi-factor authentication and forsaking insecure legacy systems are all incentives for enterprises to go passwordless.
The stigma associated with implementation expenses for passwordless systems is also slowly lifting.
The aforementioned findings appeared in a study by technology market research firm, Vanson Bourne.
Source: CSO Online
Microsoft’s new AI-powered Bing got a host of new features and upgrades since its launch earlier this year. But the latest injection is –wait for it– Ads!
But Microsoft says we should not be surprised as it confirmed the news yesterday in a blog post.
The ads will be slipped only as an experiment so far, and will be labeled as ‘sponsored’
Yet, with this, we cannot help but question how far we’ve really come from the old model of ads on search engines.
Users have started noticing the ads pop in the responses but some point out that the ads have been there since launch but not enabled for all users.
Usually, the ad sign is in a little box above companies’ names inserted in an AI-generated response to a query. So, it remains dubious as to what these companies are really advertising and how it relates to a user’s query.
Plus one response has multiple ads of different companies, all advertising different things, making it harder to parse through one objective answer.
But the most alarming is that ads like this cannot be blocked with current tools. That does not mean they can’t be blocked at all, but the paradigm of ads as predictable items in labeled locations is clearly ending.
We knew that big tech was not going to run these large language models out of charity, but maybe we expected the integration of ads to be more well-thought.
Source: TechCrunch
The U.S. has yet another proposed piece of legislation that could give the government a way to ban TikTok. But critics warn that the implications could be too wide.
Rights experts told news site Motherboard that the newly proposed RESTRICT Act contains “insanely broad language” that could lead to other apps or communications services with connections to foreign countries being banned in the U.S.
Even security tools such as virtual private networks (VPNs) that consumers widely use to encrypt and route their traffic, could be affected.
The RESTRICT Act stands for Restricting the Emergence of Security Threats that Risk Information and Communications Technology Act and is led by Senators Mark Warner and John Thune.The pair introduced the bill earlier this month.
Riana Pfefferkorn, researcher scholar at the Stanford Internet Observatory, told Motherboard in an email, “It absolutely does implicate those free speech rights for Congress to give the President the power to take ‘appropriate’ action—up to and including banning—against a particular ICTS in the name of national security or Americans’ security and safety. (Even if you trust Joe Biden with this power, would you trust Donald Trump — who tried to ban TikTok as well as WeChat while in office — with it?).”
The legislation contains vague terms like “desktop applications,” “mobile applications,” “gaming applications,” “payment applications,” and “web-based applications.”
But Rachel Cohen, communications director for Senator Warner defended the legislation saying, “This legislation is aimed squarely at companies like Kaspersky, Huawei and TikTok that create systemic risks to the United States’ national security—not at individual users”
Source: Vice
Randomness is the bane of computers but the key to good cryptography. So, Cloudflare figured out the most random solution to better encrypt your webpage: Lava lamps
Its importance is due to a process called entropy, which in cryptography means randomness or unpredictability. So encryption would typically require a high level of entropy or unpredictability to prevent attackers from detecting any type of pattern.
We could use computers but as they’re logical devices, so, they struggle with generating randomness. And that’s where the real world comes into play.
Accordingly, nothing gets more random than a picture taken of 100 lava lamps at any time of the day in different lighting conditions, in different positions, and even with people occasionally crossing in front of the camera.
Cloudflare calls it the ‘Wall of Entropy,” with lava lamps being an inherently random variable that will always change.
Cloudflare is one of the largest providers of Content Distributed Networks and is upping its encryption game with novel ideas as such.
It also has the ‘Chaotic Pendulums’ in its London office, where the movements of devices are so chaotic and “practically impossible” to predict, that Cloudflare uses readings from those devices to make long strings of numbers for key generation.
Source: XDA Developers
That’s the top tech news for today. Hashtag Trending goes to air five days a week with the daily tech news and we have a special weekend edition where we do an in depth interview with an expert on some tech development that is making the news.
Follow us on Apple, Google, Spotify or wherever you get your podcasts. Links to all the stories we’ve covered can be found in the text edition of this podcast at itworldcanada.com/podcasts
We love your comments – good or bad. You can find ITWC CIO, Jim Love on LinkedIn, Twitter, or on Mastodon as @therealjimlove on our Mastodon site technews.social. Or just leave a comment under the text version at itworldcanada.com/podcasts
I’m your host, James Roy, have a fantastic Friday!