Advertisement
Continue reading the main story
Supported by
Continue reading the main story
More than a dozen companies have popped up to offer services aimed at identifying whether photos, text and videos are made by humans or machines.
Send any friend a story
As a subscriber, you have 10 gift articles to give each month. Anyone can read what you share.
By Tiffany Hsu and Steven Lee Myers
Andrey Doronichev was alarmed last year when he saw a video on social media that appeared to show the president of Ukraine surrendering to Russia.
The video was quickly debunked as a synthetically generated deepfake, but to Mr. Doronichev, it was a worrying portent. This year, his fears crept closer to reality, as companies began competing to enhance and release artificial intelligence technology despite the havoc it could cause.
Generative A.I. is now available to anyone, and it’s increasingly capable of fooling people with text, audio, images and videos that seem to be conceived and captured by humans. The risk of societal gullibility has set off concerns about disinformation, job loss, discrimination, privacy and broad dystopia.
For entrepreneurs like Mr. Doronichev, it has also become a business opportunity. More than a dozen companies now offer tools to identify whether something was made with artificial intelligence, with names like Sensity AI (deepfake detection), Fictitious.AI (plagiarism detection) and Originality.AI (also plagiarism).
Mr. Doronichev, a Russian native, founded a company in San Francisco, Optic, to help identify synthetic or spoofed material — to be, in his words, “an airport X-ray machine for digital content.”
In March, it unveiled a website where users can check images to see if they were made by actual photographs or artificial intelligence. It is working on other services to verify video and audio.
“Content authenticity is going to become a major problem for society as a whole,” said Mr. Doronichev, who was an investor for a face-swapping app called Reface. “We’re entering the age of cheap fakes.” Since it doesn’t cost much to produce fake content, he said, it can be done at scale.
Image
The overall generative A.I. market is expected to exceed $109 billion by 2030, growing 35.6 percent a year on average until then, according to the market research firm Grand View Research. Businesses focused on detecting the technology are a growing part of the industry.
Months after being created by a Princeton University student, GPTZero claims that more than a million people have used its program to suss out computer-generated text. Reality Defender was one of 414 companies chosen from 17,000 applications to be funded by the start-up accelerator Y Combinator this winter.
Copyleaks raised $7.75 million last year in part to expand its anti-plagiarism services for schools and universities to detect artificial intelligence in students’ work. Sentinel, whose founders specialized in cybersecurity and information warfare for the British Royal Navy and the North Atlantic Treaty Organization, closed a $1.5 million seed round in 2020 that was backed in part by one of Skype’s founding engineers to help protect democracies against deepfakes and other malicious synthetic media.
Major tech companies are also involved: Intel’s FakeCatcher claims to be able to identify deepfake videos with 96 percent accuracy, in part by analyzing pixels for subtle signs of blood flow in human faces.
Within the federal government, the Defense Advanced Research Projects Agency plans to spend nearly $30 million this year to run Semantic Forensics, a program that develops algorithms to automatically detect deepfakes and determine whether they are malicious.
Even OpenAI, which turbocharged the A.I. boom when it released its ChatGPT tool late last year, is working on detection services. The company, based in San Francisco, debuted a free tool in January to help distinguish between text composed by a human and text written by artificial intelligence.
OpenAI stressed that while the tool was an improvement on past iterations, it was still “not fully reliable.” The tool correctly identified 26 percent of artificially generated text but falsely flagged 9 percent of text from humans as computer generated.
The OpenAI tool is burdened with common flaws in detection programs: It struggles with short texts and writing that is not in English. In educational settings, plagiarism-detection tools such as TurnItIn have been accused of inaccurately classifying essays written by students as being generated by chatbots.
Detection tools inherently lag behind the generative technology they are trying to detect. By the time a defense system is able to recognize the work of a new chatbot or image generator, like Google Bard or Midjourney, developers are already coming up with a new iteration that can evade that defense. The situation has been described as an arms race or a virus-antivirus relationship where one begets the other, over and over.
“When Midjourney releases Midjourney 5, my starter gun goes off, and I start working to catch up — and while I’m doing that, they’re working on Midjourney 6,” said Hany Farid, a professor of computer science at the University of California, Berkeley, who specializes in digital forensics and is also involved in the A.I. detection industry. “It’s an inherently adversarial game where as I work on the detector, somebody is building a better mousetrap, a better synthesizer.”
Despite the constant catch-up, many companies have seen demand for A.I. detection from schools and educators, said Joshua Tucker, a professor of politics at New York University and a co-director of its Center for Social Media and Politics. He questioned whether a similar market would emerge ahead of the 2024 election.
“Will we see a sort of parallel wing of these companies developing to help protect political candidates so they can know when they’re being sort of targeted by these kinds of things,” he said.
Experts said that synthetically generated video was still fairly clunky and easy to identify, but that audio cloning and image-crafting were both highly advanced. Separating real from fake will require digital forensics tactics such as reverse image searches and IP address tracking.
Available detection programs are being tested with examples that are “very different than going into the wild, where images that have been making the rounds and have gotten modified and cropped and downsized and transcoded and annotated and God knows what else has happened to them,” Mr. Farid said.
“That laundering of content makes this a hard task,” he added.
Image
The Content Authenticity Initiative, a consortium of 1,000 companies and organizations, is one group trying to make generative technology obvious from the outset. (It’s led by Adobe, with members such as The New York Times and artificial intelligence players like Stability A.I.) Rather than piece together the origin of an image or a video later in its life cycle, the group is trying to establish standards that will apply traceable credentials to digital work upon creation.
Adobe said last week that its generative technology Firefly would be integrated into Google Bard, where it will attach “nutrition labels” to the content it produces, including the date an image was made and the digital tools used to create it.
Jeff Sakasegawa, the trust and safety architect at Persona, a company that helps verify consumer identity, said the challenges raised by artificial intelligence had only begun.
“The wave is building momentum,” he said. “It’s heading toward the shore. I don’t think it’s crashed yet.”
Tiffany Hsu is a tech reporter covering misinformation and disinformation. @tiffkhsu
Steven Lee Myers covers misinformation for The Times. He has worked in Washington, Moscow, Baghdad and Beijing, where he contributed to the articles that won the Pulitzer Prize for public service in 2021. He is also the author of “The New Tsar: The Rise and Reign of Vladimir Putin.” @stevenleemyers • Facebook
A version of this article appears in print on , Section
B
, Page
1
of the New York edition
with the headline:
A.I. Spurs An Industry To Detect It. Order Reprints | Today’s Paper | Subscribe
Advertisement
Continue reading the main story
FAQs
What are 4 types of AI? ›
- Reactive machines. Reactive machines are AI systems that have no memory and are task specific, meaning that an input always delivers the same output. ...
- Limited memory. The next type of AI in its evolution is limited memory. ...
- Theory of mind. ...
- Self-awareness.
Minimizing student use of AI
Avoid the use of knowledge recognition and recall through the elimination of multiple-choice questions. Decrease the use of essays that focus on the regurgitation of knowledge from one source and that require repackaging the information as the substance of the assessment.
Artificial intelligence is generally divided into two types – narrow (or weak) AI and general AI, also known as AGI or strong AI.
What are the 5 main groups of AI? ›- Text AI.
- Visual AI.
- Interactive AI.
- Analytic AI.
- Functional AI.
In this fun one-hour class, students will learn about the Five Big Ideas in AI (Perception, Representation & Reasoning, Learning, Human-AI Interaction, and Societal Impact) through discussions and games.
Can AI manipulate humans? ›By analyzing patterns in people's online activities and social media interactions, AI algorithms can predict what a person is likely to do next. Cult leaders and dictators can use predictive models to manipulate people into doing what they want by providing incentives or punishments based on predicted behavior.
What is the smartest AI out there? ›Google DeepMind — AlphaGo
AlphaGo is considered to be one of the most intelligent AI systems in the industry due to its advanced capabilities and its ability to learn and adapt over time.
The Turing test, originally called the imitation game by Alan Turing in 1950, is a test of a machine's ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human.
What is the most advanced AI? ›OpenAI, a leading research organization in the field of artificial intelligence (AI), has recently released Chat GPT-4, the latest iteration of their language model. This release has generated a lot of excitement and anticipation, as it is the most advanced and powerful AI yet.
What type of AI is Alexa? ›Conversational AI systems are computers that people can interact with simply by having a conversation. With conversational AI, voice-enabled devices like Amazon Echo are enabling the sort of magical interactions we've dreamed of for decades.
What type of AI is Siri? ›
Siri's abilities fall under the category of narrow AI. It's not meant to be a humanoid replacement or supercomputer. Instead, it aims to be a focused assistant that can control your phone or set reminders for you.
What type of AI exist today? ›Narrow AI or weak AI: This is the type of AI that exists today. It is called narrow because it is trained to perform a single or narrow task, often far faster and better than humans can. "Weak" refers to the fact that the AI does not possess human-level, i.e., general intelligence.
What are the six artificial intelligences? ›Branches of Artificial Intelligence As AI Capabilities
There is a broad set of techniques that come in the domain of artificial intelligence such as linguistics, bias, vision, planning, robotic process automation, natural language processing, decision science, etc.
You can opt for any of 5 AI types – analytic, interactive, text, visual, and functional – or wisely combine several ones.
Are there 4 basic AI concepts? ›Most people focus on the results of AI. For those of us who like to look under the hood, there are four foundational elements to understand: categorization, classification, machine learning, and collaborative filtering.
What are the 8 foundations of AI? ›✒️ Algorithm, incompleteness theorem, computable, tractability, NP completeness, Non deterministic polynomial and probability.
What are the original 7 aspects of AI? ›The original seven aspects of AI, named by McCarthy and others at the Dartmouth Conference in 1955, include automatic computers, programming AI to use language, hypothetical neuron nets to be used to form concepts, measuring problem complexity, self-improvement, abstractions, and randomness and creativity.
What are the three 3 key elements for AI? ›The key elements of AI include: Natural language processing (NLP) Expert systems. Robotics.
What is the next big trend in AI? ›The Rise of Augmented Analytics
Every industry has found an application for augmented analytics as it transforms how organizations look at data, making it one of the definitive 2023 AI trends. Gartner predicts that by 2025, 75% of data stories will be automatically generated using augmented analytics techniques.
Even in this fictional story, ChatGPT notes that the AI was “programmed” to be malicious and evil. As with any technology, it can be used for good or for bad. The technology itself is not going to break bad on its own. AI is just like every other technology — it is just a tool that can be used for good or evil.
What are two things AI can do that humans cant? ›
AI can filter email spam, categorize and classify documents based on tags or keywords, launch or defend against missile attacks, and assist in complex medical procedures. However, if people feel that AI is unpredictable and unreliable, collaboration with this technology can be undermined by an inherent distrust of it.
Can AI create a fake person? ›Powered by artificial intelligence and deep machine learning, Fotor's AI face generator lets you create realistic human faces from scratch in seconds. Just tell Fotor what you want, such as gender, age, and other traits.
Which is the most advanced human like AI robot? ›Ameca is the world's most advanced, most realistic humanoid robot created by Engineered Arts in 2021. Ameca's first video was released publicly on Dec 1, 2021, and received a lot of attention on Twitter and TikTok.
Do any strong AI exist? ›Examples of Strong AI. Because it doesn't actually exist yet, the only true examples of AGI are found in works of science fiction like Star Trek: The Next Generation, Wall-E and Her — and most of the time they either depict a utopian version of this technology, or a dystopian one.
What is the scariest AI thought experiment? ›It's called Roko's Basilisk. This thought experiment posits that the creation of an artificial intelligence will lead to an all-powerful, future artificial intelligence that will retroactively punish anyone who did not help bring it into existence. It's been described by many as the scariest thought experiment ever.
Can AI read human minds? ›Two days ago (May 1), researchers announced the successful creation of an AI-powered decoder that's able to translate brain activity into a continuous stream of understandable language. As a result, users are able to non-invasively read another human's thoughts for the first time.
Has any AI passed the Turing test yet? ›In the decade since, many more programs have purported to pass the Turing test. Most recently, Google's AI LaMDA passed the test and even controversially convinced a Google engineer that it was “sentient.”
Is Sophia really an AI? ›Sophia is one of the most sophisticated artificial intelligence (AI) humanoid robots able to interact with people and demonstrate humanlike expressions.
Is AI self-aware? ›Technologists broadly agree that AI chatbots are not self-aware just yet, but there is some thought that we may have to re-evaluate how we talk about sentience.
Can Sophia the robot feel emotions? ›Sophia's emotional expressions were identified by the facial recognition API, as introduced in the previous section.
Does Siri count as AI? ›
Siri's evolution as an AI assistant continues as it keeps gathering data and learning from its users. A database of questions and answers makes responding to user queries easy for Siri. Siri is also adapting to cloud-based services, such as the Apple Watch.
Is Siri a human or a robot? ›I have a confession to make: I always thought that my “virtual assistant”, “Siri” was a robot—a disembodied voice created through digital technology. Siri is Apple's virtual assistant for iOS, macOS, tvOS and watchOS devices that use voice recognition and are powered by artificial intelligence (AI).
Is Alexa self aware? ›Then, the next time you contemplate hitting snooze, Alexa automatically launches into that morning routine. “The self-awareness of Alexa can make it a true assistant advisor and companion for you,” Prasad adds.
What type of AI is Jarvis? ›Jarvis, formerly known as Conversion AI, is artificial intelligence (AI) copywriting software that allows you to create content faster. Basically, instead of spending a million hours writing blog posts, ad copy, sales emails, social media posts, and catchy headlines, you can let Jarvis do most of the writing for you.
Is Siri listening all the time? ›Siri is not listening in at all, according to Apple. Instead, the capability of the software to react to a voice command is built in. Therefore, it isn't truly listened to all the time. Only a minimal quantity of audio may be stored on the iPhone, and it only starts recording after receiving the "Hey, Siri" command.
What is super AI? ›They are high-functioning systems that replicate and even surpass human intelligence but only for a specific purpose. ASI is also known as super AI or superintelligent AI.
What is the new AI that can answer anything? ›There's a new artificial intelligence-powered chatbot known as ChatGPT that can answer questions, generate essays and even write scientific papers from a short prompt.
What is the simplest form of AI? ›At its simplest form, artificial intelligence is a field, which combines computer science and robust datasets, to enable problem-solving. It also encompasses sub-fields of machine learning and deep learning, which are frequently mentioned in conjunction with artificial intelligence.
Who is the father of AI? ›Who is Hinton - father of AI. Geoffrey Hinton is an eminent figure in the field of artificial intelligence and machine learning. He holds a BA in Experimental Psychology from Cambridge and a PhD in Artificial Intelligence from Edinburgh.
What are 4 examples of machines that are smart but not AI? ›Mention four examples of machines that are smart but not AI.
Automatic gates in shopping malls / remote control drones/ a fully automatic washing machine/ Air Conditioner/ Refrigerator/ Robotic toy cars/ Television etc.
Which programming language is used for AI? ›
Python and Java are both languages that are widely used for AI. The choice between the programming languages depends on how you plan to implement AI. For example, in the case of data analysis, you would probably go with Python.
What is the new AI called? ›But it's improving rapidly. The current wave of AI products is built on a technical breakthrough called generative AI.
What are the 4 elements of AI? ›- Data Element.
- Algorithms Element.
- Platforms Element.
- Integration Element.
Generative AI steals the show
In 2023, AI is breaking out of its mold as a tool for automating routine tasks. It's diving headfirst into the world of creativity. That's right; we're talking about generative AI, the branch of AI that uses existing data to create entirely new and never before seen artifacts.
There are four types of artificial intelligence: reactive machines, limited memory, theory of mind and self-awareness.
How many main types of AI are there? ›Based on how they learn and how far they can apply their knowledge, all AI can be broken down into three capability types: Artificial narrow intelligence, artificial general intelligence and artificial superintelligence. Here's what to know about each.
What are the 6 principles of AI? ›Microsoft outlines six key principles for responsible AI: accountability, inclusiveness, reliability and safety, fairness, transparency, and privacy and security. These principles are essential to creating responsible and trustworthy AI as it moves into more mainstream products and services.
What are the seven patterns of AI? ›The seven patterns are: hyperpersonalization, autonomous systems, predictive analytics and decision support, conversational/human interactions, patterns and anomalies, recognition systems, and goal-driven systems.
What are the 4 stages of an AI workflow? ›- Digitalise and collect data. Collecting and storing data is one of the most important steps of the AI workflow. ...
- Transform and build model. ...
- Build and train. ...
- Execute. ...
- Translate the action.
AI is the ability of a machine to display human-like capabilities such as reasoning, learning, planning and creativity.
What is the most advanced AI system? ›
GPT-3 was released in 2020 and is the largest and most powerful AI model to date. It has 175 billion parameters, which is more than ten times larger than its predecessor, GPT-2.
What is the most intelligent form of AI? ›Super AI surpasses human intelligence and can perform any task better than a human.
What is the highest form of AI? ›The final type of AI is self-aware AI. This will be when machines are not only aware of emotions and mental states of others, but also their own.