The row over a manipulated photo of Rishi Sunak pulling an imperfect pint is the latest example of doctored or deepfake images attempting to disrupt politics. Using false information or imagery to alter public opinion is not new but breakthroughs in artificial intelligence threaten to take deception to a new level. Here are some recent examples of image-based disinformation. Last year a video appeared of the Ukrainian president calling on his soldiers to lay down their weapons and return to their families. It was an amateurish example of a deepfake, the term for a hoax that uses AI to create a phoney image, most commonly fake videos of people.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. Russian drones on Wednesday hit a Ukrainian port city along the border with Romania, causing significant damage and a huge fire at facilities that are key to Ukrainian grain exports. The attacks followed the end of a deal with Russia that had allowed Ukrainian shipments to world markets from the Black Sea port of Odesa. Since scrapping the deal, Russia has hammered the country's ports with strikes, compounding the blow to the key industry.

Ukraine's coastal region of Odesa was rattled by Russian drones which hit grain storage facilities in the south of the region, according to authorities in Kyiv. The grain port of Izmail, an inland port across the Danube River from NATO-member Romania, was the main target of Moscow's drone attack. "As a result of the attack, fires broke out at the facilities of the port and industrial infrastructure of the region, and an elevator was damaged," Odesa region Governor Oleh Kiper said in a statement on the Telegram messaging app. Russia's continued attacks against the Ukrainian civilian infrastructure on #Danube, in the proximity of Romania, are unacceptable. These are war crimes and they further affect UA's capacity to transfer their food products towards those in need in the world.

Tech Policy Center director Kara Fredrick explains how individuals and companies can mitigate the spread of misinformation by A.I. on'The Faulkner Focus.' British election regulators have urged politicians to pass new laws to limit spending on artificial intelligence (AI) as well as new requirements to identify AI-generated content. "The next U.K. general election is a ripe target for electronic disinformation given we are in the infancy of the AI age," Alan Mendoza, co-founder and executive director of the Henry Jackson Society, told Fox News Digital. "Many of the possible problems that may emerge have not even been considered." "As a result, we face a race against the clock to introduce appropriate protections, or run the nightmare risk of bad actors influencing campaigns and destroying public trust in our democratic process," he added.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. A Turkish drone attack killed two Kurdish local officials and their driver in northeast Syria on Tuesday in the latest such strike in the war-torn country, officials said, as talks on Syria's conflict began in Kazakhstan. The Kurdish-led authority in northeast Syria said Tuesday's strike hit a car near the town of Qamishli, killing the co-chairperson of the town's council, Yusra Darwish, and her deputy, Liman Shweish, as well as their driver. An additional local officials was wounded in the attack.

Social media platforms influence the way political campaigns are run and therefore they have become an increasingly important tool for politicians to directly interact with citizens. Previous elections in various countries have shown that social media data may significantly impact election results. In this study, we aim to predict the vote shares of parties participating in the 2023 elections in Turkey by combining social media data from various platforms together with traditional polling data. Our approach is a volume-based approach that considers the number of social media interactions rather than content. We compare several prediction models across varying time windows. Our results show that for all time windows, the ARIMAX model outperforms the other algorithms.

On the evening of Turkey's most significant elections of the past two decades, Can Semercioğlu went to bed early. For the past seven years, Semercioğlu has worked for Teyit, the largest independent fact-checking group in Turkey, but that Sunday, May 14, was surprisingly one of the quietest nights he remembers at the organization. Before the vote, opinion polls had suggested that incumbent president Recep Tayyip Erdoğan was losing support due to devastating earthquakes in southeastern Turkey that killed nearly 60,000 people and a struggling economy. However, he still managed to secure just under 50 percent of the vote. His main opponent, Kemal Kılıçdaroğlu, who heads the Millet Alliance group of opposition parties, received around 45 percent, meaning the two will face off in a second round scheduled for May 28. "That night we didn't have much work to do because people were talking about the results," Semercioğlu says.

With the rise of task-specific pre-training objectives, abstractive summarization models like PEGASUS offer appealing zero-shot performance on downstream summarization tasks. However, the performance of such unsupervised models still lags significantly behind their supervised counterparts. Similarly to the supervised setup, we notice a very high variance in quality among summary candidates from these models while only one candidate is kept as the summary output. In this paper, we propose to re-rank summary candidates in an unsupervised manner, aiming to close the performance gap between unsupervised and supervised models. Our approach improves the unsupervised PEGASUS by up to 7.27% and ChatGPT by up to 6.86% relative mean ROUGE across four widely-adopted summarization benchmarks ; and achieves relative gains of 7.51% (up to 23.73% from XSum to WikiHow) averaged over 30 zero-shot transfer setups (finetuning on a dataset, evaluating on another).

Sequence-to-sequence neural networks have recently achieved great success in abstractive summarization, especially through fine-tuning large pre-trained language models on the downstream dataset. These models are typically decoded with beam search to generate a unique summary. However, the search space is very large, and with the exposure bias, such decoding is not optimal. In this paper, we show that it is possible to directly train a second-stage model performing re-ranking on a set of summary candidates. Our mixture-of-experts SummaReranker learns to select a better candidate and consistently improves the performance of the base model. With a base PEGASUS, we push ROUGE scores by 5.44% on CNN-DailyMail (47.16 ROUGE-1), 1.31% on XSum (48.12 ROUGE-1) and 9.34% on Reddit TIFU (29.83 ROUGE-1), reaching a new state-of-the-art. Our code and checkpoints will be available at https://github.com/ntunlp/SummaReranker.

Sen. Pete Ricketts of Nebraska told Fox News Digital he's concerned about China's use of Artificial Intelligence after a report claimed pro-Chinese groups were spreading CCP propaganda using AI-generated news anchors. Turkish President Recep Tayyip Erdogan's main political opponent accused Russia of using deepfakes and other artificial intelligence (AI)-generated material to meddle in the country's upcoming presidential election. "The Russians have a vested interest in backing an Erdogan presidency to ensure that he basically stays in power, mainly because the Russians benefit [from] driving a wedge between Turkey and NATO, and they've been very successful about that in the last decade or so," Sinan Ciddi, non-resident senior fellow on Turkey at the Foundation for Defense of Democracies, told Fox News Digital. "So, in the last several days, weeks, it has been credibly reported by Turkish sources that Russian bot accounts, Twitter accounts, all sorts of disinformation campaigns have started pressing the thumb down on backing the Erdogan presidency, and that comes as no surprise." The election, scheduled for May 14 alongside parliamentary elections, has proven difficult for Erdogan as his election rival Kemal Kilicdaroglu maintains a slight lead in opinion polls.