AI is the latest tool in the cybersecurity cat-and-mouse game
Welcome to AI Decoded, Fast Company’s weekly newsletter that breaks down the most important news in the world of AI. I’m Steve Melendez, a contributing writer at Fast Company.
This week, I’m focusing on a new report that shed light on fraud in the AI era. I also look at a new proposal to create a “Manhattan Project” of AI, and offer a glimpse at our annual Next Big Things in Tech honorees.?
Sign up to receive this newsletter every week via email here .
Fraud is getting smarter, thanks to AI
AI-powered cybercrimes are on the rise, according to a recent report from cybersecurity firm Arkose Labs , letting hackers accelerate their attacks and making automated fraud more humanlike and harder to spot than ever.?
One of the best defenses, Arkose and others in the industry say, is more AI. But only about one in five enterprises reported being "very well prepared" to deploy their own AI to defend against cyberattacks, according to the report, even though 88% say they've seen "a significant increase in AI-powered bot attacks" in the past two years.
Bots are being used by hackers to take over legitimate accounts at target companies andto create numerous bogus accounts, to take advantage of sign-up bonuses, according to the report. And some bots have even begun to target large language model systems themselves. They're submitting their own prompts to AI to extract data to train their own models or even relaying prompts on behalf of their own customers, effectively selling their own subscriptions to AI services without authorization.?
Other studies have recently found that generative AI can be used to automate phishing and social-engineering attacks. A recent report from Morgan Stanley pointed out that the technology can be used to create more targeted phishing messages faster than humans, meaning the attacks can not only be launched more quickly, they can be more effective. Some AI algorithms can also make AI more adept at guessing passwords, according to the report.
Click here to read more about the rise of AI-powered cybercrime.
Commission Proposes "Manhattan Project" for AI
Among the recommendations listed in the annual report released to Congress Tuesday by the U.S.-China Economic and Security Review Commission: a call to establish "a Manhattan Project-like program" aimed at developing artificial general intelligence .
"AGI is generally defined as systems that are as good as or better than human capabilities across all cognitive domains and would surpass the sharpest human minds at every task," according to the report, which suggests Congress could provide funding for AI, data, and cloud computing companies to keep the U.S. in the lead on AI.
Whether today's technology is anywhere close to the path to artificial general intelligence is hotly debated in AI circles, but the Commission seems determined to ensure that if such superintelligent AI does pop up any time soon, it does so on U.S. soil. Commission member Jacob Helberg —a senior advisor to Palantir CEO Alex Karp, generally known for his hawkish views on China—told Reuters that China is "racing toward AGI" and suggested such steps as making it easier to build data centers could boost U.S. AI development. AI data centers have seen pushback lately amid concerns that their high energy list could raise consumer power bills, accelerate climate change, and cause other environmental issues.?
Click here to read more about the call for a Manhattan Project-like AGI program.
Looking at the next big things in AI
Fast Company released its 2024 Next Big Things in Tech honorees this week, which includes a number of companies working in artificial intelligence .?
Among those companies recognized by Fast Company is Bellwether , a company incubated at Alphabet's X research lab that processes satellite imagery with AI to help respond to disasters; Enko Chem , which uses AI to design new, environmentally friendly weedkillers for agricultural use; and Qualcomm, whose Snapdragon chips are enabling new AI features on more than 100 mobile devices from manufacturers including Samsung, Honor, and Xiaomi.
Click here to read more about the AI companies included in this year’s Next Big Things in Tech.
DeepMind's CEO is charting a path for Google AI
Google’s game plan around AI bets heavily on Gemini, a large language model that powers a raft of new features, from Google Search’s “AI Overview” summaries to Gmail’s ability to draft emails for users. To better understand Google's approach to AI, Fast Company global tech editor Harry McCracken spent time with Demis Hassabis, the CEO of Google's DeepMind division.
As Hassabis explains it, DeepMind is “the engine room of Google.” Yet DeepMind's mission is even bigger than that. Like OpenAI, Anthropic, and others, it’s working toward achieving artificial general intelligence, or AGI.
More AI coverage from Fast Company:
Want exclusive reporting and trend analysis on technology, business innovation, future of work, and design? Sign up for Fast Company Premium.