This week in AI: setting rules in the EU, defeating deepfakes
Artificial intelligence (AI) is transforming everyday life, heralding a paradigm shift beyond smarter chatbots.
Technology shows no signs of slowing down, in fact, it is accelerating.
This, as the EU’s 27 member states seem to believe that the AI ecosystem could use some speed bumps, or at least a yield mark or two.
The bloc last weekend (February 2) unanimously endorsed the final text of its AI law, overcoming concerns expressed by some countries including France, Germany and Italy about the law’s ability to hinder innovation and the status of EU-based companies. Once again in the AI arms race compared to their global peers.
Meanwhile, in the US, one venture capital firm says AI startups now account for about 80% of the pitches it sees.
Here’s a pulse check of the most important AI news tracked by PYMNTS this week.
Addressing the AI deepfakes problem
On Thursday (February 8), the Federal Communications Commission (FCC) made it illegal to use voices generated by artificial intelligence (AI) in robocalls, giving state prosecutors another tool to use against voice cloning scams, where they can prosecute the actors. The bad. Not only for the scam but also for using AI to create the voice in the robocall.
A few days ago, on Tuesday (February 6), Meta announced that it is working on detecting AI-generated content and identifying this content for its users.
Elsewhere in AI safety, Amazon and MasterCard are among dozens of members of a new government AI safety initiative. The tech giant and payments company announced on Thursday that they are part of the newly formed US Artificial Intelligence Safety Institute (AISIC) consortium.
The consortium, created by the National Institute of Standards and Technology (NIST), is designed to foster collaboration between industry and government to advance the safe use of artificial intelligence.
Regarding the use of artificial intelligence in combating fraud, the integration of artificial intelligence into traditional behavioral monitoring systems represents a leap forward in combating payment fraud. While traditional systems may excel at recognizing known patterns, they often have difficulty detecting new or complex forms of fraud. AI enhances these systems by constantly learning and adapting to emerging threats, helping organizations increase payment adoption rates while reducing fraudulent transactions.
Capture the impact of AI on enterprise workflow
Earnings calls this week from digital platforms like Snap, Roblox and Meta underscored the impact executives see AI having across user engagement and product development.
And on Friday (February 9), PYMNTS explored the role of artificial intelligence behind the recent wave of corporate layoffs across companies like UPS, BlackRock and others.
PYMNTS also reported how large AI models offered by big tech companies have served their purpose of disseminating and making revolutionary technology known across a wide global audience, and that the future of commercial applications of AI likely lies in smaller models that have fewer parameters but perform well in… Specialized tasks. .
In the market, recent moves include generative AI company Galileo launching a tool to help companies develop trustworthy AI solutions, announcing on Tuesday the launch of a new generation Retrieval Augmented (RAG) and agent analytics solution.
Google renamed its intelligent chatbot from Bard to Gemini, launched a new version of it, introduced mobile experiences and added a new subscription plan, all on Thursday.
Construction and landscaping software startup Attentive.ai has raised $7 million. The company’s AI platform helps contractors and landscapers save time and bid efficiently on contracts using automated site measurements.
Microsoft has launched an AI-focused partnership with news startup Semafor. The collaboration will see Semafor use tools from Microsoft and its partner OpenAI to develop news stories using a “global, multi-source” breaking news feed known as “Signals,” Semafor announced on Monday (February 5).
And in a little more personal news, Elon Musk He has offered financial support for a project that aims to use artificial intelligence to read ancient Roman manuscripts damaged by a volcanic eruption nearly 2,000 years ago, and announced on Wednesday (February 7) that his Musk Foundation will support the initiative.
What the experts say
The use of generative AI in areas such as decoding unstructured text, conflict resolution, automating processes, and personalizing recommendations are some of the “low-hanging fruit” where AI can have an immediate impact across B2B payments, says Ahsan Shah, Vice President Senior Head of Analytics. And Amnesty International at Billtrust, told PYMNTS.
“The first step when doing AI integration is to not walk away from the technology,” Shah added.
And when it comes to accelerating the use of generative AI – and capabilities – within the enterprise environment, “collaboration and a partnership model between computing and AI experts on one side, and domain experts in a given enterprise industry on the other side, is actually critical.” “It’s important as well as having the right compute engine and datasets under the hood,” Andy Hock, senior vice president of product and strategy at Cerebras, told PYMNTS during a conversation about the “AI Effect” series.
“I’m declaring my bias here, but I think there are very positive outcomes (it will bring AI to the business landscape), and in order to move forward with those outcomes, we must also move forward quickly with the right tools on how to do that,” Hook said. Work safely and responsibly.”
“AI will help companies move more efficiently and make better products, and it will help our human workforce and our brilliant analytical minds do the kind of work that human brains are perfectly suited to,” he said.