AI
Exclusive-AI being used for hacking and misinfo, top Canadian cyber official says
Story by By Raphael Satter •
By Raphael Satter
WASHINGTON (Reuters) - Hackers and propagandists are wielding artificial intelligence (AI) to create malicious software, draft convincing phishing emails and spread disinformation online, Canada's top cybersecurity official told Reuters, early evidence that the technological revolution sweeping Silicon Valley has also been adopted by cybercriminals.
In an interview this week, Canadian Centre for Cyber Security Head Sami Khoury said that his agency had seen AI being used "in phishing emails, or crafting emails in a more focused way, in malicious code (and) in misinformation and disinformation."
Khoury did not provide details or evidence, but his assertion that cybercriminals were already using AI adds an urgent note to the chorus of concern over the use of the emerging technology by rogue actors.
In recent months several cyber watchdog groups have published reports warning about the hypothetical risks of AI - especially the fast-advancing language processing programs known as large language models (LLMs), which draw on huge volumes of text to craft convincing-sounding dialogue, documents and more.
In March, the European police organization Europol published a report saying that models such as OpenAI's ChatGPT had made it possible "to impersonate an organisation or individual in a highly realistic manner even with only a basic grasp of the English language." The same month, Britain's National Cyber Security Centre said in a blog post that there was a risk that criminals "might use LLMs to help with cyber attacks beyond their current capabilities."
Cybersecurity researchers have demonstrated a variety of potentially malicious use cases and some now say they are beginning to see suspected AI-generated content in the wild. Last week, a former hacker said he had discovered an LLM trained on malicious material and asked it to draft a convincing attempt to trick someone into making a cash transfer.
The LLM responded with a three paragraph email asking its target for help with an urgent invoice.
"I understand this may be short notice," the LLM said, "but this payment is incredibly important and needs to be done in the next 24 hours."
Khoury said that while the use of AI to draft malicious code was still in its early stages - "there's still a way to go because it takes a lot to write a good exploit" - the concern was that AI models were evolving so quickly that it was difficult to get a handle on their malicious potential before they were released into the wild.
"Who knows what's coming around the corner," he said.
(Reporting by Raphael Satter in Washington; editing by Chris Sanders and Josie Kao)
Scientists unveil plans for underwater AI bot that detects illegal fishing
Robot powered by artificial intelligence will swim the world's seas to detect activities that harm the ocean environment
Grant and Award AnnouncementIMAGE: SCIENTISTS UNVEIL PLANS FOR NEW UNDERWATER AI BOT WHICH CAN DETECT ILLEGAL FISHING view more
CREDIT: UNIVERSITY OF SOUTHAMPTON
Scientists have started work on a new underwater artificial intelligence bot which can detect activities that harm the ocean environment.
The technology, being developed by the University of Southampton with ocean science experts RS Aqua, will be used to spot illegal fishing and protect marine mammals during offshore wind farm construction.
More than £700,000 was awarded by Innovate UK for the AI system, codenamed MARLIN, which uses underwater sensors to remotely monitor animal, human and environmental activity anywhere in the ocean before transmitting data back in real time.
Professor of Statistical Signal Processing Paul White, from the University of Southampton, said: "Using the power of artificial intelligence to monitor sound in the underwater world, combined with the ability to rapidly relay information ashore, will enable us to provide tools to protect fragile marine ecosystems and detect a range of illegal activities."
Large vessels are typically used for ocean monitoring missions – but the new MARLIN system could reduce the time ships are at sea, potentially cutting CO2 emissions by up to 75 per cent.
Dr Ryan Mowat, Research Director at RS Aqua, added: “This technology will revolutionise how we scientifically monitor our ocean environment. Currently we leave instruments underwater for months at a time and recover them before accessing their data.
“MARLIN will get that data to the internet in real time, and its implications are huge. It will help ensure that offshore construction is sensitive to marine mammal activity and will enable the monitoring of marine protected areas through the real time recognition of illegal fishing activity.”
Find out more at the MARLIN project at rsaqua.co.uk/projectmarlin.
Or read more about innovations from the University of Southampton which are addressing maritime challenges at www.southampton.ac.uk/smmi.
ENDS
283 WORDS
Fueled by new chemistry, algorithm mines fungi for useful molecules
Researchers have trained a new algorithm based on promising new targets and reinvigorated the search for clusters of genes likely to result in interesting biological compounds
Peer-Reviewed PublicationA newly described type of chemistry in fungi is both surprisingly common and likely to involve highly reactive enzymes, two traits that make the genes involved useful signposts pointing to a potential treasure trove of biological compounds with medical and chemical applications.
It was also nearly invisible to scientists until now.
In the last 15 years, the hunt for molecules from living organisms — many with promise as drugs, antimicrobial agents, chemical catalysts and even food additives — has relied on computer algorithms trained to search the DNA of bacteria, fungi and plants for genes that produce enzymes known to drive biological processes that result in interesting compounds.
“The field kind of hit a wall in the early 2000s, when the discovery process was to extract things from fungi and see what those extracts did. But we kept rediscovering the same things,” says Grant Nickles, a graduate student in the lab of Nancy Keller, professor of medical Microbiology and immunology. “As we learned more about the genes that make these cool natural products, we designed algorithms that could search for them, find targets and make the process much more efficient.”
That method also hit a wall of sorts because the algorithms only had eyes for certain types of genes.
“The main algorithms made to find natural products work great, but they are focused on genes related to three canonical backbone enzymes,” says Keller. “There have been incremental improvements to those algorithms, but you can only search the same genomes for similar genes so many times before you are, again, rediscovering the same things.”
In 2005, a community of researchers sequenced the genome of Aspergillus fumigatus, a fungus that can infect people with compromised immune systems.
“The first sequence made the hairs on my arms rise,” Keller says. “There were so many clusters of genes of the type that make these backbone enzymes that produce interesting secondary metabolites. I said, ‘Oh! There’s a lot more natural products in fungi than we ever could have guessed.’”
In subsequent research, Keller’s lab uncovered at least one cluster of genes involved in biochemical processes reliant on a backbone enzyme called isocyanide synthase, which is not one of the three “canonical” enzymes known to be common chemical workhorses across bacteria and fungi.
This month, Nickles, Keller and collaborators published a new study in the journal Nucleic Acids Research in which they describe a new algorithm they created to search fungal genomes for the groups of genes, called biosynthetic gene clusters, that synthesize isocyanide to do their work.
“I ran the new algorithm on every fungal genome that I could find on the internet — about 3,300 species — and found that that this is the fifth-largest class of natural products produce by fungi,” Nickles says. “And it was almost completely invisible prior to this study.”
More than 1,300 fungal species have clusters of genes centered on isocyanide chemistry.
“It’s so likely these gene clusters are producing something useful to the fungus, or it would be hard to explain why these genes are so common and preserved across the genomes of so many species,” says Milton Drott, co-author of the new study and a former member of Keller’s lab now working as a plant pathologist at the U.S. Department of Agriculture’s Cereal Disease Lab. “What we’ve made is an atlas of those gene clusters. You can start to see interesting patterns there that point to where to look first for significant functions.”
Highest on Keller’s list will be clusters in which the surrounding genes are ones known to tailor enzymes for different purposes or transport them to specific locations or “promoter” genes that flip the switch — on or off — for enzyme production based on conditions in their cells.
“We're looking for uniqueness,” says Keller, whose work is supported by the National Institutes of Health and who is co-founder of a company, Terra Bioforge, that makes useful natural products discovered in microbes. “Unique combinations of member genes in a cluster may tell us something about the activity of the structure. But my expectation is that we won’t be the only ones looking.”
The researchers catalogued their fungal findings on a searchable website created by co-author Brandon Oestereicher, meaning many other labs won’t even have to run an algorithmic search — a resource-heavy process that required the help of UW–Madison’s High Throughput Computing Center.
“Labs with a favorite species of fungus — that’s not unusual for people in our field, that they are focused on a species or a narrow range of species — can look their species up on the website and get enough information on the gene clusters to start their own work on isocyanides,” says Drott.
That research may reveal natural compounds with great benefits to society — antibacterial drugs, pesticides, new catalysts for industrial and pharmaceutical chemistry — but the products and purposes of this new biological chemistry are still largely unknown. Drott’s lab studies members of the fungal genus Fusarium that cause blight in grains like barley and wheat. They also have isocyanide biosynthetic gene clusters.
“This is exciting for our work, because these gene clusters may play a role in that pathogenicity, and they may provide an avenue to control the pathogen,” Drott says. “We know so little about what isocyanides can do, though, that we just don’t know what we will find. At least now we know where to start looking.”
This research was supported by grants from the National Institutes of Health (2R01GM112739-05A1 and T32 GM135066), the National Science Foundation (Graduate Research Fellowship 2137424), and the U.S. Department of Agriculture.
JOURNAL
Nucleic Acids Research
ARTICLE TITLE
Mining for a new class of fungal natural products: the evolution, diversity, and distribution of isocyanide synthase biosynthetic gene clusters
AI enables scientists to monitor impact of farming on biodiversity
UKCEH will use machine learning to identify species from photographs and recordings captured by automated monitoring stations
Business AnnouncementIMAGE: ONE OF THE NEW SOLAR-POWERED AUTOMATED MONITORING OF INSECT (AMI) TRAPS AT A FARM IN EAST NORTHAMPTONSHIRE. view more
CREDIT: UKCEH
Scientists are using automated wildlife sensors and artificial intelligence (AI) over the next four years to demonstrate the effectiveness of agri-environment and peatland restoration schemes in improving biodiversity.
The UK Centre for Ecology & Hydrology (UKCEH) will be deploying solar-powered biodiversity monitoring stations comprising camera ‘traps’ and acoustic recording equipment at farms across the country to monitor the presence of insects, birds, amphibians, bats and small mammals.
There will be stations at farms that are undertaking practices to reduce emissions, increase carbon capture and support wildlife, such as agroforestry and wildflower hay meadows. Stations will be placed in areas of the farm that do, and do not, have agri-environment measures, and will be used to measure the impacts of these schemes on species populations. They will also be located at degraded peatland areas to compare species populations on farms which remain drained for agricultural use and nearby sites that are being rewetted to provide wetland habitats that support biodiversity and absorb carbon dioxide from the atmosphere.
Researchers will then use AI software to identify species from the photographs or recordings of their calls.
The study is part of AgZero+ which is an ambitious, UKCEH-led five-year research programme supporting the UK’s transition towards domestic food production that is sustainable, carbon-neutral and has a positive effect on nature.
Professor Richard Pywell of UKCEH, who is leading the programme, explained the monitoring study would build on the institute’s long-term research which showed setting aside some land for agri-environment measures had a positive effect on biodiversity and did not affect overall crop yield.
“Mounting evidence suggests that populations of many species of insects, birds and mammals are in sharp decline in the UK and across the world, and a key driver of this change is intensification in agriculture,” he said.
“Using the latest technology, we will monitor species at farms that have different crop and land management practices, to demonstrate how a range of farming systems and agri-environment measures affect populations. Our monitoring will provide scientific evidence to inform sustainable land management policies and practices, which have a positive effect on biodiversity and mitigate climate change while enhancing crop production.”
Initially, automated stations with lighting to attract moths and high-resolution cameras to capture images for subsequent identification via AI are being set up over the summer at 10 sites, with two stations at each. The study will be at four pairs of farms, in Dorset, Wiltshire, Oxfordshire and Northamptonshire, to compare the impact of agri-environment schemes and two peatland sites, one which is being farmed and another which is being restored, in Cambridgeshire.
Moths and their caterpillars are crucial parts of ecosystems, being food sources for birds, bats and other small mammals, and amphibians, so are excellent indicators of biodiversity and quality of habitats. Moths are also important, yet underreported, pollinators of many flowers.
From next year, the stations will also have acoustic equipment to record the calls of bats, birds, amphibians and small mammals, which will then be identified via AI.
UKCEH will redeploy the automated monitoring stations at additional sites across England each year, being in operation between March and October in 2024, 2025 and 2026.
Dr Tom August, a computational ecologist at UKCEH who is overseeing the deployment of the automated biodiversity monitoring stations, explained: “New sensor and AI technology is transforming the way ecologists monitor biodiversity. Automated biodiversity monitoring stations with solar power allow us to monitor wildlife round-the-clock in remote locations without being on site, while AI technologies allow us to process the thousands of images and recordings they produce far faster than a human can.”
UKCEH will present its findings after completion of the four-year study though preliminary data will be available during the project.
Data from the study will support the biodiversity net gain strategy. From November, most developments in England must not only have no overall detrimental impact on biodiversity but enhance it by 10 per cent. Some of the peatland areas that will be studied by UKCEH will include sites being restored to compensate for damage to biodiversity due to development at nearby land.
UKCEH’s biodiversity monitoring stations will create a better understanding of how quickly populations of species respond to peatland restoration, providing important evidence for the wider biodiversity impact of biodiversity net gain.
Ends
Media queries
Photos of automated monitoring of insects (AMI) ‘traps’ at sites as part of the project are available on request. For interviews and further information, please contact Simon Williams, Media Relations Officer at UKCEH, via simwil@ceh.ac.uk or +44 (0)7920 295384.
Notes to Editors
About AgZero+
This five-year programme brings together a community of researchers and farmers to evaluate innovative farming methods and to define practical pathways to achieving “net zero plus” arable and livestock farm systems. UKCEH is leading implementation of AgZero+ and supporting its operation working in partnership with, Rothamsted Research, British Geological Survey, Plymouth Marine Laboratory, and the National Centre for Earth Observation. Funding is provided by the Natural Environment Research Council and the Biotechnology and Biological Sciences Research Council, which are part of UK Research and Innovation.
About the UK Centre for Ecology & Hydrology
The UK Centre for Ecology & Hydrology (UKCEH) is a world-leading centre for excellence in environmental sciences across water, land and air. Our research makes a major contribution to the development of sustainable, productive farming systems that are resilient to climate change and protect biodiversity. We identify key drivers of biodiversity change, develop tools and technologies for monitoring biodiversity, and provide robust socio-economic and environmental solutions for restoring biodiversity.
The UK Centre for Ecology & Hydrology is a strategic delivery partner for the Natural Environment Research Council.
ceh.ac.uk / Twitter: @UK_CEH / LinkedIn: UK Centre for Ecology & Hydrology
METHOD OF RESEARCH
Observational study
SUBJECT OF RESEARCH
Animals
Call for papers: Theme issue on artificial intelligence (AI) and human factors—towards successful application of AI in health care
IMAGE: CALL FOR PAPERS: THEME ISSUE ON ARTIFICIAL INTELLIGENCE (AI) AND HUMAN FACTORS—TOWARDS SUCCESSFUL APPLICATION OF AI IN HEALTH CARE view more
CREDIT: JMIR PUBLICATIONS
JMIR Human Factors (2023 Impact Factor 2.7) Editor-in-Chief: Andre Kushniruk, BA, MSc, PhD, FACMI and Guest Editor: Elizabeth Borycki RN, PhD, FIAHIS, FACMI, FCAHS welcome submissions to a special theme issue examining "Artificial Intelligence (AI) and Human Factors—Towards Successful Application of AI in Health Care."
This issue will focus on identifying the types of applications, users, and contexts where AI has proven to lead to successful adoption. In addition, there is also a need to identify specific issues related to the broad field of human factors, where numerous challenges have impacted the adoption of AI technologies in health care. Identifying key factors associated with the successful adoption of AI technology will become increasingly critical as innovative AI applications emerge and become more widely deployed.
Considerable advances in research and applications of artificial intelligence (AI) in health care have been made over the past several decades. The contexts of use have diversified, as the focus of AI has shifted from applications for acute care to use in health promotion and applications in the home. However, despite these advances, a range of challenges and human factors issues need to be considered to achieve the full potential of AI applications in clinical and health care settings.
Submissions may include, but are not limited to, the following areas of research:
- Human factors of AI applications
- Usability of AI applications
- Policy issues in implementing AI in health care
- Regulatory issues in implementing AI in health care
- Ethical issues related to the design and implementation of AI in health care (including data privacy)
- Safety issues with AI in health care
- Usability or end-user experience with AI
- AI system adoption in health care
- Cognition and AI (including automation bias and fairness)
- Socio-technical aspects of AI in health care
- Human-centered design and evaluation of AI systems
We welcome original research papers, short communications, viewpoints, and reviews that provide insight into the intersection of AI in health care and human factors.
The deadline for submissions is October 15, 2023. All accepted manuscripts will be published as part of the JMIR Human Factors special theme issue on “AI and Human Factors.”
To learn more please visit: https://humanfactors.jmir.org/announcements/394
###
About JMIR Publications
JMIR Publications is a leading, born-digital, open access publisher of 30+ academic journals and other innovative scientific communication products that focus on the intersection of health and technology. Its flagship journal, the Journal of Medical Internet Research, is the leading digital health journal globally in content breadth and visibility, and it is the largest journal in the medical informatics field.
To learn more about JMIR Publications, please visit https://www.JMIRPublications.com or connect with us via Twitter, LinkedIn, YouTube, Facebook, and Instagram.
Head office: 130 Queens Quay East, Unit 1100, Toronto, ON, M5A 0P6 Canada
Media contact: communications@JMIR.org
No comments:
Post a Comment